Data Over Cable Service Interface Specifications Physical Layer Specificaction

Data Over Cable Service Interface Specifications
DOCSIS
3.0
Physical Layer Specification
CM-SP-PHYv3.0-I10-111117
ISSUED
Notice
This DOCSIS specification is the result of a cooperative effort
undertaken at the direction of Cable Television Laboratories, Inc. for the
benefit of the cable industry and its customers. This document may
contain references to other documents not owned or controlled by
CableLabs. Use and understanding of this document may require
access to such other documents. Designing, manufacturing,
distributing, using, selling, or servicing products, or providing services,
based on this document may require intellectual property licenses from
third parties for technology referenced in this document.
Neither CableLabs nor any member company is responsible to any
party for any liability of any nature whatsoever resulting from or arising
out of use or reliance upon this document, or any document referenced
herein. This document is furnished on an "AS IS" basis and neither
CableLabs nor its members provides any representation or warranty,
express or implied, regarding the accuracy, completeness,
noninfringement, or fitness for a particular purpose of this document, or
any document referenced herein.
2006 - 2011 Cable Television Laboratories, Inc.
All rights reserved.
CM-SP-PHYv3.0-I10-111117 Data Over Cable Service Interface Specifications
ii CableLabs
11/17/11
Document Status Sheet
Document Control Number CM-SP-PHYv3.0-I10-111117
Document Title Physical Layer Specification
Revision History I01 Released 08/04/06
I02 Released 12/22/06
Date November 17, 2011
Status Work in
Progress
Draft Issued Closed
Distribution Restrictions Author Only CL/Member CL/ Member/
Vendor
Public

Key to Document Status Codes
Work in Progress An incomplete document, designed to guide discussion and generate feedback
that may include several alternative requirements for consideration.
Draft A document in specification format considered largely complete, but lacking review
by Members and vendors. Drafts are susceptible to substantial change during the
review process.
Issued A stable document, which has undergone rigorous member and vendor review and
is suitable for product design and development, cross-vendor interoperability, and
for certification testing.
Closed A static document, reviewed, tested, validated, and closed to further engineering
change requests to the specification through CableLabs.

Trademarks
CableCARD, CableHome, CableLabs, CableNET, CableOffice, CablePC, DCAS, DOCSIS,
DPoE, EBIF, eDOCSIS, EuroDOCSIS, EuroPacketCable, Go2Broadband
SM
, M-Card, M-CMTS,
OCAP, OpenCable, PacketCable, PCMM, PeerConnect, and tru2way are marks of Cable Television
Laboratories, Inc. All other marks are the property of their respective owners.
Physical Layer Specification CM-SP-PHYv3.0-I10-111117
11/17/11 CableLabs
iii
Contents
1 SCOPE .................................................................................................................................................................. 1
1.1 Introduction and Purpose ............................................................................................................................... 1
1.2 Background .................................................................................................................................................... 1
1.2.1 Broadband Access Network ................................................................................................................... 1
1.2.2 Network and System Architecture .......................................................................................................... 2
1.2.3 Service Goals ......................................................................................................................................... 3
1.2.4 Statement of Compatibility ..................................................................................................................... 3
1.2.5 Reference Architecture .......................................................................................................................... 4
1.2.6 DOCSIS 3.0 Documents ......................................................................................................................... 4
1.3 Requirements ................................................................................................................................................. 5
1.4 Conventions ................................................................................................................................................... 5
1.5 Organization of Document ............................................................................................................................. 6
2 REFERENCES .................................................................................................................................................... 7
2.1 Normative References .................................................................................................................................... 7
2.2 Informative References .................................................................................................................................. 8
2.3 Reference Acquisition .................................................................................................................................... 8
3 TERMS AND DEFINITIONS ............................................................................................................................ 9
4 ABBREVIATIONS AND ACRONYMS .......................................................................................................... 16
5 FUNCTIONAL ASSUMPTIONS ..................................................................................................................... 19
5.1 Equipment Assumptions .............................................................................................................................. 19
5.1.1 Frequency Plan .................................................................................................................................... 19
5.1.2 Compatibility with Other Services ....................................................................................................... 19
5.1.3 Fault Isolation Impact on Other Users ................................................................................................ 19
5.1.4 Cable System Terminal Devices ........................................................................................................... 20
5.2 RF Channel Assumptions ............................................................................................................................. 20
5.2.1 Transmission Downstream ................................................................................................................... 20
5.2.2 Transmission Upstream ....................................................................................................................... 21
5.3 Transmission Levels .................................................................................................................................... 22
5.4 Frequency Inversion ..................................................................................................................................... 22
6 PHYSICAL MEDIA DEPENDENT SUBLAYER SPECIFICATION .......................................................... 23
6.1 Scope ........................................................................................................................................................... 23
6.2 Upstream ...................................................................................................................................................... 23
6.2.1 Overview .............................................................................................................................................. 23
6.2.2 Signal Processing Requirements.......................................................................................................... 25
6.2.3 Modulation Formats ............................................................................................................................ 27
6.2.4 R-S Encode .......................................................................................................................................... 28
6.2.5 Upstream R-S Frame Structure for DOCSIS 3.0 Multiple Transmit Channel mode Enabled ............. 28
6.2.6 Upstream R-S Frame Structure for DOCSIS 3.0 Multiple Transmit Channel mode Not Enabled ...... 30
6.2.7 TDMA Byte Interleaver ........................................................................................................................ 30
6.2.8 Scrambler (Randomizer) ...................................................................................................................... 33
6.2.9 TCM Encoder ....................................................................................................................................... 34
6.2.10 Preamble Prepend ............................................................................................................................... 37
6.2.11 Modulation Rates ................................................................................................................................. 38
6.2.12 S-CDMA Framer and Interleaver ........................................................................................................ 38
6.2.13 S-CDMA Framer.................................................................................................................................. 45
iv CableLabs
11/17/11
6.2.14 Symbol Mapping .................................................................................................................................. 48
6.2.15 S-CDMA Spreader ............................................................................................................................... 55
6.2.16 Transmit Pre-Equalizer ....................................................................................................................... 61
6.2.17 Spectral Shaping .................................................................................................................................. 63
6.2.18 Relative Processing Delays.................................................................................................................. 65
6.2.19 Transmit Power Requirements ............................................................................................................. 65
6.2.20 Burst Profiles ....................................................................................................................................... 76
6.2.21 Burst Timing Convention ..................................................................................................................... 82
6.2.22 Fidelity Requirements .......................................................................................................................... 83
6.2.23 Upstream Demodulator Input Power Characteristics ......................................................................... 92
6.2.24 Upstream Electrical Output from the CM ............................................................................................ 93
6.2.25 Upstream CM Transmitter Capabilities ............................................................................................... 94
6.3 Downstream ................................................................................................................................................. 95
6.3.1 Downstream Protocol and Interleaving Support ................................................................................. 95
6.3.2 Downstream Electrical Input to CM .................................................................................................... 95
6.3.3 CM BER Performance ......................................................................................................................... 96
6.3.4 Downstream Multiple Receiver Capabilities ....................................................................................... 98
6.3.5 Non-Synchronous DS Channel Support ............................................................................................. 101
ANNEX A TIMING REQUIREMENTS FOR SUPPORTING BUSINESS SERVICES OVER
DOCSIS ........................................................................................................................................................... 103
A.1 CMTS ........................................................................................................................................................ 103
A.2 CM ............................................................................................................................................................. 103
ANNEX B ADDITIONS AND MODIFICATIONS FOR EUROPEAN SPECIFICATION ........................ 104
B.1 Scope ......................................................................................................................................................... 104
B.1.1 Introduction and Purpose .................................................................................................................. 104
B.1.2 Background ........................................................................................................................................ 104
B.1.3 Requirements ..................................................................................................................................... 105
B.1.4 Conventions ....................................................................................................................................... 105
B.1.5 Organization of Document................................................................................................................. 105
B.2 References .................................................................................................................................................. 105
B.2.1 Normative References ........................................................................................................................ 105
B.2.2 Informative References ...................................................................................................................... 106
B.2.3 Reference Acquisition ........................................................................................................................ 106
B.3 Terms and Definitions ................................................................................................................................ 106
B.4 Abbreviations and Acronyms ..................................................................................................................... 106
B.5 Functional Assumptions ............................................................................................................................. 106
B.5.1 Equipment Assumptions ..................................................................................................................... 106
B.5.2 RF Channel Assumptions ................................................................................................................... 107
B.5.3 Transmission Levels ........................................................................................................................... 109
B.5.4 Frequency Inversion .......................................................................................................................... 109
B.6 Physical Media Dependent Sublayer Specification .................................................................................... 109
B.6.1 Scope .................................................................................................................................................. 109
B.6.2 Upstream............................................................................................................................................ 109
B.6.3 Downstream ....................................................................................................................................... 140
ANNEX C MPEG HEADER SYNCHRONIZATION AND RECOVERY ................................................... 145
C.1 MPEG Header Synchronization and Recovery in the North American Technology Option...................... 145
C.2 MPEG Header Synchronization and Recovery in the European Technology Option ................................ 145
APPENDIX I EXAMPLE PREAMBLE SEQUENCE ..................................................................................... 146
I.1 Introduction................................................................................................................................................ 146
I.2 Example Preamble Sequence ..................................................................................................................... 146
11/17/11 CableLabs
v
APPENDIX II S-CDMA FRAMING ............................................................................................................... 148
II.1 Coded Subsymbol Numbering ................................................................................................................... 148
II.2 Uncoded Subsymbol Numbering ............................................................................................................... 148
II.3 Framer Output Numbering ......................................................................................................................... 149
II.4 Comments .................................................................................................................................................. 149
APPENDIX III AMBIENT TEMPERATURE AND WIND LOADING EFFECTS .................................... 150
III.1 Synchronization Tolerances to Plant Delay Variations .............................................................................. 150
III.2 Change in Propagation Delay Due to Temperature Changes ..................................................................... 151
III.2.1 Fiber Delay Changes Due to Temperature ........................................................................................ 151
III.2.2 Coaxial Cable Delay Changes Due to Temperature.......................................................................... 152
III.2.3 Delay Change Due to Wind ............................................................................................................... 152
APPENDIX IV DESCRIPTION OF UPSTREAM TRANSMIT CHANNEL SET CAPABILITY:
EXAMPLE CALCULATIONS FOR REPORTING AND FIGURING THE NUMBER OF ACTIVE
CHANNELS SUPPORTED .................................................................................................................................... 153
APPENDIX V DESCRIPTION OF UPSTREAM CHANNEL POWER CONTROL WITH
MULTIPLE UPSTREAM CHANNELS ................................................................................................................ 155
V.1 DOCSIS 2.0 Parameters Extended to Multiple Transmit Channel mode ................................................... 155
V.2 New Parameters in DOCSIS 3.0 Upstream Power Control ("loading," P
load_n
, P
load_min_set
, Dynamic
Range Window, P
low_multi_n
) .................................................................................................................................... 155
V.3 Example Upstream Power Control with Multiple Transmit Channel mode Enabled ................................. 156
V.4 Examples Regarding Concurrent and Consecutive Changes in P
r_n
and P
load_min_set
................................... 158
APPENDIX VI EXAMPLE SPURIOUS EMISSIONS NOISE POWER LIMITS WITH
MULTIPLE CHANNELS BURSTING ................................................................................................................. 160
APPENDIX VII ACKNOWLEDGEMENTS ..................................................................................................... 161
APPENDIX VIII REVISION HISTORY ........................................................................................................... 162
VIII.1 Engineering Changes incorporated into CM-SP-PHYv3.0-I02-061222 ................................................ 162
VIII.8 Engineering Change incorporated into CM-SP-PHYv3.0-I09-101008 .................................................. 163

vi CableLabs
11/17/11
Figures
Figure 11 - The DOCSIS Network .............................................................................................................................. 2
Figure 12 - Transparent IP Traffic Through the Data-Over-Cable System .................................................................. 3
Figure 13 - Data-over-Cable Reference Architecture .................................................................................................. 4
Figure 61 - Upstream Signal-Processing Sequence ................................................................................................... 25
Figure 62 - TDMA Upstream Transmission Processing ............................................................................................ 26
Figure 63 - S-CDMA Upstream Transmission Processing ........................................................................................ 27
Figure 64 - Example Frame Structures with Flexible Burst Length Mode, DOCSIS 3.0 Operation .......................... 29
Figure 65 - Byte Interleaver Operation ...................................................................................................................... 32
Figure 66 - Interleaver Operation for Last Interleaver Block (with Shortened Last Codeword)................................ 32
Figure 67 - T Mode Calculations ............................................................................................................................... 33
Figure 68 - Scrambler Structure ................................................................................................................................. 34
Figure 69 - Convolutional encoder ............................................................................................................................ 34
Figure 610 - Repetitive Patterns of Byte Mapping to Symbol Map Bits for TCM .................................................... 35
Figure 611 - Example Byte to Bit Assignment for 64-QAM ..................................................................................... 36
Figure 612 - Example of return to zero bits followed by "0" ..................................................................................... 37
Figure 613 - Timestamp Snapshot ............................................................................................................................. 39
Figure 614 - Mini-slot Mapping with Two Codes per mini-slot, 128 Active Codes .................................................. 41
Figure 615 - Mini-slot Mapping with Three Codes per mini-slot, 126 Active Codes ................................................ 42
Figure 616 - Mini-slot Mapping with Four Codes per Mini-slot, 124 Active Codes, Codes 0, 1, 5, and 125
Unused; Selectable Active Codes Mode 2 ........................................................................................................... 42
Figure 617 - S-CDMA Spreader-on and Spreader-off Intervals ................................................................................ 44
Figure 618 - Subframe structure ................................................................................................................................ 46
Figure 619 - Symbol Numbering With and Without TCM ........................................................................................ 48
Figure 620 - Symbol Constellations ........................................................................................................................... 51
Figure 621 - QPSK Gray and Differential Symbol Mapping ..................................................................................... 52
Figure 622 - 8-QAM Symbol Mapping ..................................................................................................................... 52
Figure 623 - 16-QAM Symbol Mapping ................................................................................................................... 53
Figure 624 - 32-QAM Symbol Mapping .................................................................................................................. 53
Figure 625 - 64-QAM Symbol Mapping ................................................................................................................... 54
Figure 626 - QPSK and 8-QAM TCM Symbol Mapping .......................................................................................... 54
Figure 627 - 16-QAM and 32-QAM TCM Symbol Mapping .................................................................................... 55
Figure 628 - 64-QAM and 128-QAM TCM Symbol Mapping ................................................................................. 55
Figure 629 - Code Hopping Random Number Generator .......................................................................................... 60
Figure 630 - Transmit Pre-Equalizer Structure .......................................................................................................... 62
Figure 631 - Nominal TDMA Burst Timing .............................................................................................................. 82
Figure 632 - Worst-Case TDMA Burst Timing ......................................................................................................... 83
Figure 633 - 6 MHz DOCSIS Standard Receive Channel Profile 01 ...................................................................... 100
Figure 634 - Example Manufacturer Receive Channel Profile ................................................................................ 101
Figure B1 - Transmit Pre-Equalizer Structure ......................................................................................................... 115
Figure B2 - 8 MHz DOCSIS Standard Receive Channel Profile CLAB-8M-004 ................................................... 144
11/17/11 CableLabs
vii
Tables
Table 11 - DOCSIS 3.0 Series of Specifications ......................................................................................................... 4
Table 12 - DOCSIS 3.0 Related Specifications ........................................................................................................... 5
Table 51 - Assumed Downstream RF Channel Transmission Characteristics ........................................................... 20
Table 52 - Assumed Upstream RF Channel Transmission Characteristics ................................................................ 21
Table 61 - Burst Size ................................................................................................................................................. 29
Table 62 - Interleaver Operating Parameters ............................................................................................................. 31
Table 63 - I/Q Mapping ............................................................................................................................................. 49
Table 64 - Definition of Differential Quadrant Coding ............................................................................................. 50
Table 65 - Maximum Channel Width for Operation with a DOCSIS 3.0 CMTS ...................................................... 64
Table 66 - Maximum Channel Width for Operation with a pre-3.0-DOCSIS CMTS ................................................ 64
Table 67 - Constellation Gains and Per Channel Power Limits with One Channel in the Transmit Channel
Set ........................................................................................................................................................................ 69
Table 68 - Constellation Gains and Per Channel Power Limits with Two Channels in the Transmit
Channel Set .......................................................................................................................................................... 69
Table 69 - Constellation Gains and Per Channel Power Limits with Three or Four Channels in the
Transmit Channel Set ........................................................................................................................................... 69
Table 610 - Minimum Transmit Power P
min
for Multiple Transmit Channel mode Not Enabled .............................. 76
Table 611 - Burst Profile Attributes .......................................................................................................................... 76
Table 612 - User Unique Burst Parameters ............................................................................................................... 77
Table 613 - Timing Offset for Modulation Rate Changes ......................................................................................... 81
Table 614 - Spurious Emissions ................................................................................................................................ 85
Table 615 - Adjacent Channel Spurious Emissions Requirements Relative to the Per Channel Transmitted Burst
Power Level for Each Channel ............................................................................................................................ 86
Table 616 - Adjacent Channel Spurious Emissions Goals Relative to the Per Channel Transmitted Burst Power
Level for Each Channel ....................................................................................................................................... 87
Table 617 - Spurious Emissions Requirements in the Upstream Frequency Range Relative to the Per
Channel Transmitted Burst Power Level for Each Channel................................................................................. 87
Table 618 - Spurious Emissions Goals in the Upstream Frequency Range Relative to the Per Channel
Transmitted Burst Power Level for Each Channel ............................................................................................... 87
Table 619 - Single Channel Filter Amplitude Distortion ........................................................................................... 91
Table 620 - Upstream Channel Demodulator Input Power Characteristics ............................................................... 93
Table 621 - Electrical Output from CM ..................................................................................................................... 93
Table 622 - Electrical Input to CM ............................................................................................................................ 96
Table B1 - DOCSIS 3.0 Series of Specifications..................................................................................................... 105
Table B2 - Assumed Downstream RF Channel Transmission Characteristics ......................................................... 107
Table B3 - Assumed Upstream RF Channel Transmission Characteristics ............................................................. 108
Table B4 - Constellation Gains and Per Channel Power Limits with One Channel in the Transmit
Channel Set ........................................................................................................................................................ 120
Table B5 - Constellation Gains and Per Channel Power Limits with Two Channels in the Transmit
Channel Set ........................................................................................................................................................ 121
Table B6 - Constellation Gains and Per Channel Power Limits with Three or Four Channels in the
Transmit Channel Set ......................................................................................................................................... 121
Table B7 - Minimum Transmit Power P
min
for Multiple Transmit Channel mode Not Enabled ............................. 127
Table B8 - Burst Profile Attributes .......................................................................................................................... 128
viii CableLabs
11/17/11
Table B9 - User Unique Burst Parameters ............................................................................................................... 129
Table B10 - Spurious Emissions .............................................................................................................................. 133
Table B11 - Adjacent Channel Spurious Emissions Requirements Relative to the Per Channel
Transmitted Burst Power Level for Each Channel ............................................................................................. 135
Table B12 - Adjacent Channel Spurious Emissions Goals Relative to the Per Channel Transmitted Burst Power
Level for Each Channel ..................................................................................................................................... 135
Table B13 - Spurious Emissions Requirements in the Upstream Operating Frequency Range Relative to
the Per Channel Transmitted Burst Power Level for Each Channel .................................................................. 135
Table B14 - Spurious Emissions Goals in the Upstream Operating Frequency Range Relative to the Per
Channel Transmitted Burst Power Level for Each Channel............................................................................... 136
Table B15 - Single Channel Electrical Output from CM ......................................................................................... 139
Table B16 - Electrical Input to CM ......................................................................................................................... 141
Table III1 - Allowable Plant Timing Drift ............................................................................................................... 150
11/17/11 CableLabs
1
1 SCOPE
1.1 Introduction and Purpose
This specification is part of the DOCSIS
family of specifications developed by Cable Television Laboratories

(CableLabs). In particular, this specification is part of a series of specifications that defines the third generation of
high-speed data-over-cable systems. This specification was developed for the benefit of the cable industry, and
includes contributions by operators and vendors from North America, Europe, and other regions.
There are differences in the cable spectrum planning practices adopted for different networks in the world.
Therefore, two options for physical layer technology are included, which have equal priority and are not required to
be interoperable. One technology option is based on the downstream multi-program television distribution that is
deployed in North America using 6 MHz channeling. The other technology option is based on the corresponding
European multi-program television distribution. Both options have the same status, notwithstanding that the
document structure does not reflect this equal priority. The first of these options is defined in Sections 5 and 6,
whereas the second is defined by replacing the content of those sections with the content of Annex B.
Correspondingly, [ITU-T J.83-B] and [CEA-542-B] apply only to the first option, and [EN 300 429] only to the
second. Compliance with this document requires compliance with the one or the other of these implementations, not
with both. It is not required that equipment built to one option shall interoperate with equipment built to the other.
These optional physical-layer technologies allow operators flexibility in mandated areas of operation, including any
frequency planning, EMC (electromagnetic compatibility), and safety requirements. For example, the 6 MHz
downstream based option defined in Sections 5 and 6 might be deployable within an 8 MHz channel plan.
Compliance with frequency planning and EMC requirements is not covered by this specification and remains the
operators' responsibility. In this respect, [FCC15] and [FCC76] are relevant to North America and [EG 201 212],
[EN 50083-1], [EN 50083-2], [EN 50083-7], [EN 50083-10], [EN 60950-1], [EN 61000-6-1], and [EN 61000-6-3]
are relevant to the European Union.
Backwards compatibility with earlier versions of that technology [DOCSIS RFI 2.0] is only ensured within the same
technology options referred to above and not between the two options.
1.2 Background
1.2.1 Broadband Access Network
A coaxial-based broadband access network is assumed. This may take the form of either an all-coax or hybrid-
fiber/coax (HFC) network. The generic term "cable network" is used here to cover all cases.
A cable network uses a tree-and-branch architecture with analog transmission. The key functional characteristics
assumed in this document are the following:
Two-way transmission.
A maximum optical/electrical spacing between the CMTS and the most distant CM of 100 miles in each
direction, although typical maximum separation may be 10-15 miles.
A maximum differential optical/electrical spacing between the CMTS and the closest and most distant modems
of 100 miles in each direction, although this would typically be limited to 15 miles.
At a propagation velocity in fiber of approximately 1.5 ns/ft, 100 miles of fiber in each direction results in a round-
trip delay of approximately 1.6 ms.
2 CableLabs
11/17/11
1.2.2 Network and System Architecture
1.2.2.1 The DOCSIS Network
The elements that participate in the provisioning of DOCSIS services are shown in the following figure:
Provisioning
Systems
IPv6
CPE
IPv4
CPE
Back Office Network HFC Network Home Network
CMTS
CM
IPv6
CPE
IPv4
CPE
CM
NMS
HFC

Figure 11 - The DOCSIS Network

The CM connects to the operator's HFC network and to a home network, bridging packets between them. Many
CPEs' devices can connect to the CMs' LAN interfaces. CPE devices can be embedded with the CM in a single
device, or they can be separate standalone devices, as shown in Figure 11. CPE devices may use IPv4, IPv6 or both
forms of IP addressing. Examples of typical CPE devices are home routers, set-top devices, personal computers, etc.
The CMTS connects the operator's back office and core network with the HFC network. Its main function is to
forward packets between these two domains, and between upstream and downstream channels on the HFC network.
Various applications are used to provide back office configuration and other support to the devices on the DOCSIS
network. These applications use IPv4 and/or IPv6 as appropriate to the particular operator's deployment. The
following applications include:
Provisioning Systems:
1

The DHCP servers provide the CM with initial configuration information, including the device IP address(es),
when the CM boots.
The Config File server is used to download configuration files to CMs when they boot. Configuration files are in
binary format and permit the configuration of the CM's parameters.
The Software Download server is used to download software upgrades to the CM.
The Time Protocol server provides Time Protocol clients, typically CMs, with the current time of day.
Certificate Revocation server provides certificate status.

1
Section modified per PHYv3.0-N-07.0599-1 on 1/25/08 by KN.
11/17/11 CableLabs
3
NMS:
The SNMP Manager allows the operator to configure and monitor SNMP Agents, typically the CM and the
CMTS.
The Syslog server collects messages pertaining to the operation of devices.
The IPDR Collector server allows the operator to collect bulk statistics in an efficient manner.
1.2.3 Service Goals
As cable operators have widely deployed high-speed data services on cable television systems, the demand for
bandwidth has increased. Additionally, networks have scaled to such a degree that IPv4 address constraints are
becoming a burden on network operations. To this end, CableLabs' member companies have decided to add new
features to the DOCSIS specification for the purpose of increasing channel capacity, enhancing network security,
expanding addressability of network elements, and deploying new service offerings.
The DOCSIS system allows transparent bi-directional transfer of Internet Protocol (IP) traffic, between the cable
system head-end and customer locations, over an all-coaxial or hybrid-fiber/coax (HFC) cable network. This is
shown in simplified form in Figure 12.

Figure 12 - Transparent IP Traffic Through the Data-Over-Cable System

1.2.4 Statement of Compatibility
This document specifies an interface, commonly referred to as DOCSIS 3.0, which is the third generation of the
interface, commonly referred to as DOCSIS 1.x and 2.0. DOCSIS 3.0 MUST be backward- and forward-compatible
with equipment built to the previous specifications. DOCSIS 3.0-compliant CMs MUST interoperate seamlessly with
DOCSIS 2.0 and DOCSIS 1.x CMTSs, albeit in the 2.0 and 1.x modes, as the case may be. DOCSIS 3.0-compliant
CMTSs MUST seamlessly support DOCSIS 2.0 and DOCSIS 1.x CMs.
4 CableLabs
11/17/11
1.2.5 Reference Architecture
Rx
Rx
Tx
Tx
Fiber
Node Fiber
Node
M-CMTS
Core
EQAM
Upstream
Receiver
DOCSIS
TimingServer
WideArea
Network
NetworkSide
Interface(NSI)
OperationsSupport
SystemsInterface
(OSSI)
CableModem
toCPE
Interface
(CMCI)
Downstream
External-Phy
Interface(DEPI)
DOCSISTiming
Interface(DTI)
EdgeResource
Management
Interfaces
(ERMI)
Downstream
RFInterface
(DRFI)
Cable
Modem
(CM)
Operations
SupportSystem
Edge
Resource
Manager
PhysicalLayer
Interface
(PHY)
Downstream
RF
Network
Upstream
RF
Network
Opt.
Tx
Opt.
Rx
Fiber
Node
M-CMTS
I-CMTS
DistributionHuborHeadend
Coax
Distribution
Fiber
Distribution
NOTE:Lightershadedareasare
relatedfunctionality,butoutofthe
scopeofthisdocument.
Customer
Premises
Equipment
MAC&UpperLayer
ProtocolsInterface
(MULPI)&
SecurityInterface
(SEC)

Figure 13 - Data-over-Cable Reference Architecture

The reference architecture for data-over-cable services and interfaces is shown in Figure 13.
1.2.6 DOCSIS 3.0 Documents
A list of the specifications in the DOCSIS 3.0 series is provided in Table 11. For further information, please refer to
http://www.cablemodem.com.
Table 11 - DOCSIS 3.0 Series of Specifications
Designation Title
CM-SP-PHYv3.0 Physical Layer Specification
CM-SP-MULPIv3.0 Media Access Control and Upper Layer Protocols Interface Specification
CM-SP-OSSIv3.0 Operations Support System Interface Specification
CM-SP-SECv3.0 Security Specification
CM-SP-CMCIv3.0 Cable Modem CPE Interface Specification
2

This specification defines the interface for the physical layer.
Related DOCSIS specifications are listed in Table 12.

2
This row and table following added per PHYv3.0-N-08.0702-2 on 12/23/08 by CP.
11/17/11 CableLabs
5
Table 12 - DOCSIS 3.0 Related Specifications
Designation Title
CM-SP-eDOCSIS eDOCSIS Specification
CM-SP-DRFI Downstream Radio Frequency Interface Specification
CM-SP-DTI DOCSIS Timing Interface Specification
CM-SP-DEPI Downstream External PHY Interface Specification
CM-SP-DSG DOCSIS Set-Top Gateway Interface Specification
CM-SP-ERMI Edge Resource Manager Interface Specification
CM-SP-M-OSSI M-CMTS Operations Support System Interface Specification
CM-SP-L2VPN Layer 2 Virtual Private Networks Specification
CM-SP-TEI TDM Emulation Interfaces Specification

1.3 Requirements
Throughout this document, the words that are used to define the significance of particular requirements are
capitalized. These words are:
"MUST" This word means that the item is an absolute requirement of this specification.
"MUST NOT" This phrase means that the item is an absolute prohibition of this specification.
"SHOULD" This word means that there may exist valid reasons in particular circumstances to ignore this
item, but the full implications should be understood and the case carefully weighed before
choosing a different course.
"SHOULD NOT" This phrase means that there may exist valid reasons in particular circumstances when the
listed behavior is acceptable or even useful, but the full implications should be understood and
the case carefully weighed before implementing any behavior described with this label.
"MAY" This word means that this item is truly optional. One vendor may choose to include the item
because a particular marketplace requires it or because it enhances the product, for example;
another vendor may omit the same item.

This document defines many features and parameters, and a valid range for each parameter is usually specified.
Equipment (CM and CMTS) requirements are always explicitly stated. Equipment must comply with all mandatory
(MUST and MUST NOT) requirements to be considered compliant with this specification. Support of non-
mandatory features and parameter values is optional.
1.4 Conventions
In this specification the following convention applies any time a bit field is displayed in a figure. The bit field should
be interpreted by reading the figure from left to right, then, top to bottom, with the MSB being the first bit read and
the LSB being the last bit read.
6 CableLabs
11/17/11
1.5 Organization of Document
Section 1 provides an overview of the DOCSIS 3.0 series of specifications including the DOCSIS reference
architecture and statement of compatibility.
Section 2 includes a list of normative and informative references used within this specification.
Section 3 defines the terms used throughout this specification.
Section 4 defines the acronyms used throughout this specification.
Section 5 provides a technical overview and lists the DOCSIS 3.0 key features for the functional area of this
specification.
Section 6 defines the interface requirements and the performance requirements for the CM downstream and upstream
physical layer, and for the CMTS upstream physical layer.
Annex A describes the timing requirements for the CM and CMTS for supporting business services with cable
systems using DOCSIS.
Annex B contains the PHY requirements for the European technology option for DOCSIS.
Annex C contains the MPEG header synchronization and recovery requirements.
Appendix I presents an illustrative example of the DOCSIS upstream programmable preamble superstring.
Appendix II presents an algorithmic description of the subsymbol mapping for S-CDMA framing.
Appendix III describes the impact of temperature and wind loading on timing variation of the signaling across a
cable plant, and the tolerances and impacts on the DOCSIS communications system to such variations.
Appendix IV contains example calculations for reporting from the CM to the CMTS the number of upstream active
channels of various bandwidths supported by the CM.
Appendix V provides an explanation of the power control algorithm with multiple upstream channels.
Appendix VI provides examples illustrating the calculation of the CM noise power limits for the upstream when
more than one channel is bursting.
Appendix VII contains acknowledgements to contributors of this specification.
Appendix VIII includes the engineering change history of this specification.
11/17/11 CableLabs
7
2 REFERENCES
2.1 Normative References
In order to claim compliance with this specification, it is necessary to conform to the following standards and other
works as indicated, in addition to the other requirements of this specification. Notwithstanding, intellectual property
rights may be required to use or implement such normative references.
[CEA-542-B] CEA-542-B: CEA Standard: Cable Television Channel Identification Plan," July 2003.
[DRFI] Downstream Radio Frequency Interface Specification, CM-SP-DRFI-I12-111117,
November 17, 2011, Cable Television Laboratories, Inc.
[DTI] DOCSIS Timing Interface, CM-SP-DTI-I05-081209, December 9, 2008, Cable Television
Laboratories, Inc.
[EG 201 212] ETSI EG 201 212 V1.2.1: Electrical safety; Classification of interfaces for equipment to be
connected to telecommunication networks, November 1998.
[EN 300 429] ETSI EN 300 429 V1.2.1: Digital Video Broadcasting (DVB); Framing structure, channel
coding and modulation for cable systems, April 1998.
[EN 50083-1] CENELEC EN 50083-1: Cable networks for television signals, sound signals and
interactive services -- Part 1: Safety requirements, 2002.
interactive services -- Part 10: System performance for return paths, March 2002.
interactive services -- Part 2: Electromagnetic compatibility for equipment, 2005.
interactive services -- Part 7: System performance, April 1996.
[EN 60950-1] CENELEC EN 60950-1: Information technology equipment - Safety -- Part 1: General
requirements, December 2001.
[EN 61000-6-1] CENELEC EN 61000-6-4: Electromagnetic compatibility (EMC) -- Part 6-1: Generic
standards - Immunity for residential, commercial and light-industrial environments, October
2001.
[EN 61000-6-3] CENELEC EN 61000-6-3: Electromagnetic compatibility (EMC) -- Part 6-3: Generic
standards - Emission standard for residential, commercial and light-industrial environments,
2003.
[FCC15] Code of Federal Regulations, Title 47, Part 15, October 2005.
[FCC76] Code of Federal Regulations, Title 47, Part 76, October 2005.
[ISO 13818] ISO/IEC 13818-1, Information Technology Generic Coding Of Moving Pictures And
Associated Audio Systems Recommendation H.222.0, February 2000.
[ISO/IEC-61169-
24]
ISO/IEC-61169-24, Radio-frequency connectors - Part 24: Sectional specification - Radio
frequency coaxial connectors with screw coupling, typically for use in 75 ohm cable
distribution systems (type F), 2001.
[ITU-T J.83-B] Annex B to ITU-T Rec. J.83 (4/97), Digital multi-program systems for television sound and
data services for cable distribution.
[MULPI] Media Access Control and Upper Layer Protocols Interface Specification, CM-SP-
MULPIv3.0-I17-111117, November 17, 2011, Cable Television Laboratories, Inc.
[OSSI3.0] DOCSIS 3.0 Operations Support System Interface Specification, CM-SP-OSSIv3.0-I16-
111117, November 17, 2011, Cable Television Laboratories, Inc.
8 CableLabs
11/17/11
[DOCSIS RFI 2.0] Data-Over-Cable Service Interface Specifications, Radio Frequency Interface Specification
v2.0, CM-SP-RFIv2.0-C02-090422, April 22, 2009, Cable Television Laboratories, Inc.
[SCTE 02] ANSI/SCTE 02, Specification for "F" Port, Female Indoor, 2006.
[TEI] Business Services over DOCSIS, TDM Emulation Interface Specification, CM-SP-TEI-I06-
100611, June 11, 2010, Cable Television Laboratories, Inc.

2.2 Informative References
[CableLabs1] Digital Transmission Characterization of Cable Television Systems, Cable Television
Laboratories, Inc., November 1994. URL:
http://www.cablelabs.com/downloads/digital_transmission.pdf
[NCTA] NCTA Recommended Practices for measurements on Cable Television Systems National
Cable Television Association, Washington DC, 2
nd
Edition, revised October 1993.

2.3 Reference Acquisition
Cable Television Laboratories, Inc., http://www.cablelabs.com/
CENELEC: European Committee for Electro-technical Standardization, http://www.cenelec.org
EIA: Electronic Industries Alliance, http://www.eia.org/new_contact/
ETSI: European Telecommunications Standards Institute,
http://www.etsi.org/services_products/freestandard/home.htm
Internet Engineering Task Force (IETF), http://www.ietf.org/html/
ISO: International Organization for Standardization (ISO), http://www.iso.org/iso/en/xsite/contact/contact.html
ITU: International Telecommunications Union (ITU), http://www.itu.int/home/contact/index.html
11/17/11 CableLabs
9
3 TERMS AND DEFINITIONS
3

This specification uses the following terms:
4

Active Codes The set of spreading codes which carry information in an S-CDMA upstream.
The complementary set, the unused codes, are idle and are not transmitted.
Reducing the number of active codes below the maximum value of 128 may
provide advantages including more robust operation in the presence of colored
noise.
Allocation A group of contiguous mini-slots in a MAP which constitutes a single transmit
opportunity.
Availability In cable television systems, availability is the long-term ratio of the actual RF
channel operation time to scheduled RF channel operation time (expressed as a
percent value) and is based on a bit error rate (BER) assumption.
Bandwidth Allocation Map
(MAP)
The MAC Management Message that the CMTS uses to allocate transmission
opportunities to cable modems (MAP).
Bit Error Rate (BER) The percentage of bits that have errors relative to the total number of bits
received in a transmission, usually expressed as ten to a negative power.
Burst A single continuous RF signal from the upstream transmitter, from transmitter
on to transmitter off.
Cable Modem (CM) A modulator-demodulator at subscriber locations intended for use in conveying
data communications on a cable television system. (CM).
Cable Modem Termination
System (CMTS)
Cable modem termination system, located at the cable television system head-
end or distribution hub, which provides complementary functionality to the
cable modems to enable data connectivity to a wide-area network.
Capture Bandwidth (CBW) The sum of the Tuning Bands in the TB List in MHz.
Carrier Hum Modulation The peak-to-peak magnitude of the amplitude distortion relative to the RF
carrier signal level due to the fundamental and low-order harmonics of the
power-supply frequency.
Carrier-to-Noise Ratio (C/N)
(CNR)
The ratio of signal power to noise power in the defined measurement
bandwidth. For digital modulation, CNR = E
s
/N
o
, the energy-per-symbol to
noise-density ratio; the signal power is measured in the occupied bandwidth,
and the noise power is normalized to the modulation-rate bandwidth. For
video, the measurement bandwidth is 4 MHz (C/N).
Channel (See RF Channel.)
Channel Bonding A logical process that combines the data packets received on multiple
independent channels into one higher-speed data stream. Channel bonding can
be implemented independently on upstream channels or downstream channels.
Chip Each of the 128 bits comprising the S-CDMA spreading codes.
Chip Rate The rate at which individual chips of the S-CDMA spreading codes are
transmitted. (1280 to 5120 kHz). E
s
/N
o
.
Codeword An element of an error-correcting code used to detect and correct transmission
errors.

3
Two definitions added to this table per PHYv3.0-N-0366-2 by kn on 2/8/07. Definition revised in this table per PHYv3.0-N-
07.0456-1 # 1 on 6/12/07 by KN; revised per PHYv3.0-N-011.1020-2 and PHYv3.0-N-11.1021-1 on 11/4/11 by PO.
4
Replaced this entire table per PHYv3.0-N-06.0328-2 by GO on 1/11/07.
10 CableLabs
11/17/11
Codeword Error Rate The ratio of the number of uncorrectable code words to the total number of
code words sent without errors, with corrected errors and with uncorrectable
errors.
Composite Second Order Beat
(CSO)
The peak of the average level of distortion products due to second order
nonlinearities in cable system equipment.
Composite Triple Beat (CTB) The peak of the average level of distortion components due to third-order
nonlinearities in cable system equipment.
Cross-modulation A form of television signal distortion where modulation from one or more
television channels is imposed on another channel or channels.
Customer Premises Equipment
(CPE)
Equipment at the end user's premises; may be provided by the end user or the
service provider.
Decibel-Millivolt (dBmV) A dB measurement system wherein 0 dBmV is defined as 1 millivolt over 75
ohms.
Decibels (dB) A unit to measure the relative levels of current, voltage or power. An increase
of 3 dB indicates a doubling of power, an increase of 10 dB indicates a 10x
increase in power, and an increase of 20 dB indicates a 100x increase in power.
Demodulator Module A physical entity in the CM that demodulates a block of one or more
contiguous channels of a single bandwidth (6 MHz or 8 MHz) within the
output from a single tuner.
Distribution Hub A location in a cable television network which performs the functions of a
head-end for customers in its immediate area, and which receives some or all
of its television program material from a Master Head-end in the same
metropolitan or regional area.
DOCSIS 1.x Abbreviation for "DOCSIS 1.0 or 1.1." DOCSIS stands for Data-Over-Cable
Service Interface Specifications.
DOCSIS 2.0 Mode A CM operates in this mode when: 1) Multiple Transmit Channel (MTC)
Mode is disabled; 2) the Enable 2.0 Mode configuration setting in the REG-
RSP is set to 1 (Enable) explicitly or by default; and 3) it operates on at least
one upstream channel using the burst descriptors associated with IUC 9, 10,
and 11 as opposed to IUC 5 and 6. A CM is enabled for DOCSIS 2.0 Mode
when the Enable 2.0 Mode configuration setting in the REG-RSP is set to 1
(Enable). A CM may be enabled for DOCSIS 2.0 Mode but may not be
operating in DOCSIS 2.0 Mode. When a CM has MTC Mode enabled, the CM
is not considered to be in DOCSIS 2.0 Mode even if some of the upstream
channels it is using are operating with post-1.1 DOCSIS physical layer
mechanisms. Therefore, "DOCSIS 2.0 Mode" does not have relevance for a
CM operating in MTC Mode.
Downstream In cable television, the direction of transmission from the head-end to the
subscriber.
Downstream Channel Physical layer characteristics and MAC layer parameters and functions
associated to a DOCSIS forward channel.
Dynamic Host Configuration
Protocol (DHCP)
An Internet protocol used for assigning network-layer (IP) addresses.
Dynamic Range The ratio between the greatest signal power that can be transmitted over a
multichannel analog transmission system without exceeding distortion or other
performance limits, and the least signal power that can be utilized without
exceeding noise, error rate or other performance limits.
Dynamic Range Window
(DRW)
A 12 dB range defining the maximum power difference between multiple
transmitters in a CM in Multiple Transmit Channel mode.
11/17/11 CableLabs
11
Electronic Industries Alliance
(EIA)
A voluntary body of manufacturers which, among other activities, prepares and
publishes standards.
Extended Upstream Frequency
Range
An optional upstream frequency range over which a CM may be capable of
transmitting. This is defined to be 5-85 MHz.
F Connector (F conn) A male F-connector is the final piece of hardware (familiar to subscribers) on a
drop cable. It is cylindrical with a center pin sticking out that plugs into the
female F-connector on a set-top box, cable ready TV or VCR.
Floor A mathematical function that returns the highest-valued integer that is less than
or equal to a given value.
Forward Channel The direction of RF signal flow away from the head-end toward the end user;
equivalent to Downstream.
Forward Error Correction
(FEC)
FEC enables the receiver to detect and fix errors to packets without the need
for the transmitter to retransmit packets.
Frame See MAC frame, S-CDMA frame, and MPEG frame.
Frequency Division Multiple
Access (FDMA)
A multiple access technology that separates users by putting each traffic
channel on a discrete frequency band.
Group Delay The difference in transmission time between the highest and lowest of several
frequencies through a device, circuit or system.
Guard Band Minimum time, measured in modulation symbols, allocated between bursts in
the upstream referenced from the symbol center of the last symbol of a burst to
the symbol center of the first symbol of the following burst. The guard band
should be at least the duration of five symbols plus the maximum system timing
error. The guard band should be at least the duration of five symbols plus the
maximum system timing error.
Guard Time Guard time, measured in modulation symbols, is similar to the guard band,
except that it is measured from the end of the last symbol of one burst to the
beginning of the first symbol of the preamble of an immediately following
burst. Thus, the guard time is equal to the guard band 1.
Harmonic Related Carrier
(HRC)
A method of spacing television channels on a cable television system in exact 6
MHz increments, with all carrier frequencies harmonically related to a common
reference.
Head-end The central location on the cable network that is responsible for injecting
broadcast video and other signals in the downstream direction. See also Master
Head-End, Distribution Hub.
Header Protocol control information located at the beginning of a protocol data unit.
Hertz (Hz) A unit of frequency equivalent to one cycle per second. See also kilohertz
(kHz) and megahertz (MHz).
Hum Modulation Undesired modulation of the television visual carrier by the fundamental or
low-order harmonics of the power supply frequency or other low-frequency
disturbances.
Hybrid Fiber/Coaxial System
(HFC)
A broadband bidirectional shared-media transmission system using fiber trunks
between the head-end and the fiber nodes, and coaxial distribution from the
fiber nodes to the customer locations.
Impulse Noise Noise characterized by non-overlapping transient disturbances.
Incremental Related Carriers
(IRC)
A method of spacing NTSC television channels on a cable television system in
which all channels except 5 and 6 correspond to the standard channel plan,
used to reduce composite triple beat distortions.
12 CableLabs
11/17/11
Information Element The fields that make up a MAP and define individual grants, deferred grants,
etc.
International Electrotechnical
Commission (IEC)
An international standards body.
International Organization for
Standardization (ISO)
An international standards body, commonly known as the International
Standards Organization.
Internet Engineering Task
Force (IETF)
A body responsible, among other things, for developing standards used in the
Internet.
Internet Protocol (IP) The computer network protocol (analogous to written and verbal languages)
that all machines on the Internet must know so that they can communicate with
one another. IP is a layer 3 (network layer) protocol in the OSI model. The vast
majority of IP devices today support IP version 4 (IPv4) defined in RFC-791,
although support for IP version 6 (IPv6, RFC-2460) is increasing.
Internet Protocol Detail
Record Service
The record formatter and exporter functions of the CMTS that creates the data
record compliant to the IPDR/BSR based on the DOCSIS schemas.
Interval Usage Code (IUC) A field in MAPs and UCDs to link burst profiles to grants.
Jitter The fluctuation in the arrival time of a regularly scheduled event such as a
clock edge or a packet in a stream of packets. Jitter is defined as fluctuations
above 10 Hz.
Latency The time taken for a signal element to pass through a device.
Layer A subdivision of the Open System Interconnection (OSI) architecture,
constituted by subsystems of the same rank.
Local Area Network (LAN) A non-public data network in which serial transmission is used for direct data
communication among data stations located on the user's premises.
Logical (Upstream) Channel A MAC entity identified by a unique channel ID and for which bandwidth is
allocated by an associated MAP message. A physical upstream channel may
support multiple logical upstream channels. The associated UCD and MAP
messages completely describe the logical channel.
MAC Frame MAC header plus optional protocol data unit.
Maximum Downstream
Bonded Channels (MDBC)
Maximum number of downstream bonded channels supported by the cable
modem. (See Channel Bonding).
Media Access Control (MAC) The MAC sublayer is the part of the data link layer that supports topology-
dependent functions and uses the services of the Physical Layer to provide
services to the Logical Link Control (LLC) sublayer. It can also be a
component of a networking software stack. In the OSI 7-layer model, the
Media Access Control is a part of layer 2, the data link layer.
Megahertz (MHz) One million cycles per second.
Micro-reflections Echoes in the forward or reverse transmission path due to impedance
mismatches between the physical plant components. Micro-reflections are
distinguished from discrete echoes by having a time difference (between the
main signal and the echo) on the order of one microsecond. Micro-reflections
cause departures from ideal amplitude and phase characteristics for the
transmission channel.
Microsecond (s) One millionth of a second.
Microvolt (V) One millionth of a volt.
Millisecond (ms) One thousandth of a second.
Millivolt (mV) One thousandth of a volt.
11/17/11 CableLabs
13
Mini-slot A mini-slot is an integer multiple of 6.25 s increments.
Modulation Error Ratio
(MER)
MER measures the cluster variance in dB caused by the transmit waveform. It
includes the effects of ISI, spurious, phase noise, and all other transmitter
degradations.
Modulation Rate The signaling rate of the upstream modulator (1280 to 5120 kHz). In S-CDMA
it is the chip rate. In TDMA, it is the channel symbol rate.
Nanosecond One millionth of a second.
National Cable
Telecommunications
Association (NCTA)
A voluntary association of cable television operators which, among other
things, provides guidance on measurements and objectives for cable television
systems in the USA.
National Television Systems
Committee (NTSC)
Committee which defined the analog color television broadcast standard used
today in North America.
Number of Allocated Codes The total number of codes which a single CM uses in a single S-CDMA frame.
This number is determined by the size of the grants in mini-slots and the
mapping of these mini-slots to S-CDMA frames (note that a CM may receive
multiple grants which are mapped to a single S-CDMA frame). The number of
allocated codes can be in the range of the number of Codes per Mini-slot to the
number of active codes, and may vary from frame to frame, but is constant
within an S-CDMA frame.
Phase Noise Rapid, short-term, random fluctuations in the phase of a wave, caused by time
domain instabilities.
Physical Layer Layer 1 in the Open System Interconnection (OSI) architecture; the layer that
provides services to transmit bits or groups of bits over a transmission link
between open systems and which entails electrical, mechanical and
handshaking procedures (PHY).
Physical Media Dependent
Sublayer (PMD)
A sublayer of the Physical Layer which is concerned with transmitting bits or
groups of bits over particular types of transmission link between open systems
and which entails electrical, mechanical and handshaking procedures (PMD).
Picosecond (ps) One trillionth of a second.
Primary Channel See Primary Downstream Channel.
Primary Downstream Channel The downstream channel from which a CM derives CMTS master clock
timing for upstream transmission. All other concurrently received channels are
called "secondary downstream channels."
Protocol A set of rules and formats that determines the communication behavior of layer
entities in the performance of the layer functions.
Quadrature Amplitude
Modulation (QAM)
A method of modulating digital signals onto a radio-frequency carrier signal
involving both amplitude and phase coding.
Quadrature Phase Shift Keying
(QPSK)
A method of modulating digital signals onto a radio-frequency carrier signal
using four phase states to code two digital bits.
Radio Frequency (RF) In cable television systems, electromagnetic signals in the range 5 to
1000 MHz (RF).
Radio Frequency Channel
(RFC)
The frequency spectrum occupied by a signal. Usually specified by center
frequency and bandwidth parameters.
Return Loss The parameter describing the attenuation of a guided wave signal (e.g., via a
coaxial cable) returned to a source by a device or medium resulting from
reflections of the signal generated by the source.
Reverse Channel The direction of signal flow towards the head-end, away from the subscriber;
equivalent to Upstream.
14 CableLabs
11/17/11
Root Mean Square (RMS) A mathematical method of computing an "average" magnitude of n elements by
taking the square root of the mean of the elements squared.
S-CDMA Frame A two dimensional representation of mini-slots, where the dimensions are
codes and time. An S-CDMA frame is composed of p active codes in the code
dimension and K spreading intervals in the time dimension. Within the S-
CDMA frame, the number of mini-slots is determined by the number of codes
per mini-slot (c) and p, the number of active codes in the S-CDMA frame.
Each S-CDMA frame thus contains s mini-slots, where s=p/c, and each
mini-slot contains c*K information (QAM) symbols.
S-CDMA Subframe A subframe is a vertically-smaller subset of an S-CDMA frame over which
interleaving is performed, where the vertical dimension is R' codes, where R'
p (the number of active codes). A subframe is generally used to constrain the
interleaving region to be of a similar size to the Reed-Solomon codeword in
order to provide protection from impulse noise.
Selectable Active Codes (SAC) A methodology to determine the set of active codes and its complement, the set
of unused codes. In SAC mode 1, a consecutive set of codes starting with code
0 are unused. In SAC mode 2, the active codes are selectable via a 128-bit
string.
Service Identifier (SID) A Service Flow Identifier assigned by the CMTS (in addition to a Service Flow
Identifier) to an Active or Admitted Upstream Service Flow. [14 bits] (SID).
Spread Symbol At the output of the spreader, a group of 128 chips which comprises a single S-
CDMA spreading code, and is the result of spreading a single information
(QAM constellation) symbol, which is referred to as a "spread symbol."
Spreader-Off S-CDMA Burst A transmission from a single CM in a spreader-off frame on an S-CDMA
channel defined by the time in which the cable modem's transmitter turns on to
the time it turns off. There will generally be several spreader off bursts in a
spreader-off frame.
Spreader-Off S-CDMA Frame TDMA mini-slots on an S-CDMA channel in which the spreader is turned off.
These are differentiated from TDMA bursts on a TDMA channel in that, for
example, the number of mini-slots per spreader-off S-CDMA burst frame is
constrained to be the same as the number of mini-slots in a spreader-on S-
CDMA frame (s). This number of mini-slots will be less than the number of
TDMA mini-slots in a TDMA channel over the same time interval if the
number of active codes is significantly less than 128.
Spreading Codes A family of orthogonal digital code words used in S-CDMA direct-sequence
spread-spectrum modulation.
Spreading Interval The period of a spread symbol (128 chips) is called a "spreading interval."
Standard Upstream Frequency
Range
The required upstream frequency range over which a CM is to be capable of
transmitting. In the technology option that uses 6 MHz downstream
channelization, this is 5-42 MHz. In the technology option that uses 8 MHz
downstream channelization, this is 5-65 MHz.
Sublayer A subdivision of a layer in the Open System Interconnection (OSI) reference
model.
Subscriber (See End User.)
Synchronous-Code Division
Multiple Access
(S-CDMA)
A multiple access physical layer technology in which different transmitters can
share a channel simultaneously. The individual transmissions are kept distinct
by assigning each transmission an orthogonal "code." Orthogonality is
maintained by all transmitters being precisely synchronized with one another.
11/17/11 CableLabs
15
Tick 6.25 s time intervals that are the reference for upstream mini-slot definition
and upstream transmission times.
Time Division Multiple Access
(TDMA)
A digital technology that enables a large number of users to access, in
sequence, a single radio frequency channel without interference by allocating
unique time slots to each user within each channel.
Transit Delay The time difference between the instant at which the first bit of a PDU crosses
one designated boundary, and the instant at which the last bit of the same PDU
crosses a second designated boundary.
Tuner Module A physical entity in the CM that converts a block of one or more contiguous
channels of a single bandwidth (6 MHz or 8 MHz) from the RF input from the
cable plant to an intermediate frequency suitable for distribution to a
Demodulator Module.
Tuning Band (TB) Defines a single continuous frequency interval, in MHz, located anywhere in
the downstream band. In the technology option that uses 6 MHz
channelization, the downstream band is 108 MHz to 870 MHz, with a CM
option of tuning downstream 108 MHz to 1002 MHz. In the technology option
that uses 8 MHz channelization, the downstream band is 108 MHz to 862
MHz, with a CM option of tuning downstream 108 MHz to 1006 MHz.
Tuning Band List (TB List) A list of one or more Tuning Bands supported by the CM that defines the cable
modem tuning capabilities.
Type/Length/Value (TLV) An encoding of three fields, in which the first field indicates the type of
element, the second the length of the element, and the third field the value of
the element.
Upstream The direction from the subscriber location toward the head-end.
Upstream Channel Physical layer characteristics and MAC layer parameters and functions
associated to a DOCSIS reverse channel.
Upstream Channel Descriptor
(UCD)
The MAC Management Message used to communicate the characteristics of
the upstream physical layer to the cable modems.

16 CableLabs
11/17/11
4 ABBREVIATIONS AND ACRONYMS
This specification uses the following abbreviations:
5

ANSI American National Standards Institute
AWGN Additive White Gaussian Noise
BER Bit Error Rate
CableLabs Cable Television Laboratories
CBW Capture Bandwidth
CENELEC European Committee for Electrotechnical Standardization
CL CableLabs
CM Cable Modem
CMTS Cable Modem Termination System
CNR Carrier-to-Noise Ratio
C/N Carrier-to-Noise Ratio
CPE Customer Premises Equipment
CRC Cyclic Redundancy Check
CSO Composite Second Order Composite Second Order Beat
CTB Composite Triple Beat
CW Continuous Wave
dB Decibel
DBC-REQ Dynamic Bonding Change Request MAC Message
dBc Decibels relative to carrier power
DHCP Dynamic Host Configuration Protocol
DOCSIS Data-Over-Cable Service Interface Specifications
DOCSIS 1.x Data-Over-Cable Service Interface Specifications version 1.0 or 1.1
DRFI Downstream Radio Frequency Interface
DRW Dynamic Range Window
DS Downstream
DTI DOCSIS Timing Interface
EC Errors Corrected
EIA Electronic Industries Association
ETSI European Telecommunications Standards Institute
EU Errors Uncorrectable
FCC Federal Communications Commission
FDMA Frequency Division Multiple Access
FEC Forward Error Correction
FC Frame Control
FM Frequency Modulation
GF Galois Field

5
Replaced this entire table per PHYv3.0-N-06.0328-2 by GO on 1/11/07.
11/17/11 CableLabs
17
HFC Hybrid Fiber-Coaxial
HRC Harmonic Related Carrier
I In-phase modulation component
ICMP Internet Control Message Protocol
IE Information Element
IEC International Electrotechnical Commission
IETF Internet Engineering Task Force
IP Internet Protocol
IPDR Internet Protocol Detail Record
IPv4 Internet Protocol version 4
IPv6 Internet Protocol version 6
IRC Incremental Related Carrier(s)
ISO International Standards Organization
ITU International Telecommunications Union
ITU-T Telecommunication Standardization Sector of the International Telecommunication Union
IUC Interval Usage Code
LAN Local Area Network
LFSR Linear Feedback Shift Register
LLC Logical Link Control
LSB Least Significant Bit
MAC Media Access Control
M-CMTS Modular Cable Modem Termination System
MDBC Maximum Downstream Bonded Channels
MDD MAC Domain Descriptor
MER Modulation Error Ratio
MIB Management Information Base
M/N Relationship of integer numbers M,N that represents the ratio of the downstream symbol clock
rate to the DOCSIS master clock rate
MPEG Moving Picture Experts Group
MSB Most Significant Bit
MSC Maximum Scheduled Codes
MTC Multiple Transmit Channel
MULPI MAC and Upper Layer Protocols Interface
NA Number of active codes
NACO Network Access Control Object
NCTA National Cable & Telecommunications Association
NMS Network Management System
ns Nanosecond
NTSC National Television Systems Committee
OCAP Open Cable Application Platform
OSI Open Systems Interconnection
18 CableLabs
11/17/11
PAL Phase Alternating Line
PAR Peak to Average Ratio
PDU Protocol Data Unit
PHY Physical Layer
PLL Phase-Locked Loop
PMD Physical Media Dependent sublayer
PRS Primary Reference Source
Q Quadrature modulation component
QAM Quadrature Amplitude Modulation
QPSK Quadrature Phase Shift Keying
RCC Receive Channel Configuration
RCP Receive Channel Profile
REG-REQ Registration Request MAC Message
RF Radio Frequency
RFC Request For Comments
RFI Radio Frequency Interface
RM Receive Module
RMS Root Mean Square
RNG-RSP Ranging Response MAC Message
RS Symbol rate
SAC Selectable Active Codes
S-CDMA Synchronous Code Division Multiple Access
SCTE Society of Cable Telecommunications Engineers
SECAM Squentiel couleur mmoire (Sequential color with memory)
SID Service Identifier
SNMP Simple Network Management Protocol
STD Standard Channel Plan
TB Tuning Band
TCS Transmit Channel Set
TDM Time Division Multiplexing
TDMA Time Division Multiple Access
TE-CMTS Time Division Multiplexing Emulation Cable Modem Termination System
TEI TDM Emulation Interface
TLV Type/Length/Value
TV Television
TCM Trellis Code Modulation
TSP Time Division Multiplexing Service Processor
UCD Upstream Channel Descriptor
UGS Unsolicited Grant Service
XOR Exclusive Or

11/17/11 CableLabs
19
5 FUNCTIONAL ASSUMPTIONS
This section describes the characteristics of a cable television plant, assumed to be for the purpose of operating a
data-over-cable system. It is not a description of CMTS or CM parameters. The data-over-cable system MUST be
interoperable within the environment described in this section.
Whenever a reference in this section to frequency plans, or compatibility with other services, conflicts with any legal
requirement for the area of operation, the latter shall take precedence. Any reference to NTSC analog signals in 6
MHz channels does not imply that such signals are physically present.
5.1 Equipment Assumptions
5.1.1 Frequency Plan
In the downstream direction, the cable system is assumed to have a pass band with a lower edge of either 54 MHz or
108 MHz, and an upper edge that is implementation-dependent but is typically in the range of 300 to 1002 MHz.
Within that pass band, NTSC analog television signals in 6 MHz channels are assumed present on the standard, HRC
or IRC frequency plans of [CEA-542-B], as well as other narrowband and wideband digital signals.
In the upstream direction, the cable system may have a 5-30 MHz, 5-42 MHz or 5-85 MHz pass band. NTSC analog
television signals in 6 MHz channels may be present, as well as other signals.
5.1.2 Compatibility with Other Services
The CM and CMTS MUST coexist with any services on the cable network.
In particular,
CM and CMTS MUST be interoperable in the cable spectrum assigned for CMTS and CM interoperation while
the balance of the cable spectrum is occupied by any combination of television and other signals; and
CM and CMTS MUST NOT cause harmful interference to any other services that are assigned to the cable
network in spectrum outside of that allocated to the CMTS.
Harmful interference is understood as:
No measurable degradation (highest level of compatibility),
No degradation below the perceptible level of impairments for all services (standard or medium level of
compatibility), or
No degradation below the minimal standards accepted by the industry (for example, FCC for analog video
services) or other service provider (minimal level of compatibility).
5.1.3 Fault Isolation Impact on Other Users
As CMTS transmissions are on a shared-media, point-to-multipoint system, fault-isolation procedures should take
into account the potential harmful impact of faults and fault-isolation procedures on numerous users of the data-over-
cable, video and other services.
For the interpretation of harmful impact, see Section 5.1.2 above.
20 CableLabs
11/17/11
5.1.4 Cable System Terminal Devices
The CM MUST meet and preferably exceed all applicable regulations for Cable System Termination Devices and
Cable Ready Consumer Equipment as defined in FCC Part 15 [FCC15] and Part 76 [FCC76]. None of these specific
requirements may be used to relax any of the specifications contained elsewhere within this document.
5.2 RF Channel Assumptions
The data-over-cable system, configured with at least one set of defined physical-layer parameters (e.g., modulation,
interleaver depth, etc.) from the range of configurable settings described in this specification, MUST be
interoperable on cable networks having characteristics defined in this section. This is accomplished in such a manner
that the forward error correction provides for equivalent operation in a cable system both with and without the
impaired channel characteristics, described below.
5.2.1 Transmission Downstream
The RF channel transmission characteristics of the cable network in the downstream direction are described in Table
51. These numbers assume total average power of a digital signal in a 6 MHz channel bandwidth for carrier levels
unless indicated otherwise. For impairment levels, the numbers in Table 51 assume average power in a bandwidth
in which the impairment levels are measured in a standard manner for a cable TV system. For analog signal levels,
the numbers in Table 51 assume peak envelope power in a 6 MHz channel bandwidth. All conditions are present
concurrently. No combination of the following parameters will exceed any stated interface limit defined elsewhere in
this specification.
Table 51 - Assumed Downstream RF Channel Transmission Characteristics
Parameter Value
Frequency range Cable system normal downstream operating range is from
50 MHz to 1002 MHz. However, the values in this table
apply only at frequencies 108 MHz (including Pre-3.0
DOCSIS modes).
RF channel spacing (design bandwidth) 6 MHz
Transit delay from head-end to most distant customer 0.800 ms (typically much less)
Carrier-to-noise ratio in a 6 MHz band Not less than 35 dB1,2
Carrier-to-Composite triple beat distortion ratio Not less than 41 dB1,2
Carrier-to-Composite second order distortion ratio Not less than 41 dB1,2
Carrier-to-Cross-modulation ratio Not less than 41 dB1,2
Carrier-to-any other discrete interference (ingress) Not less than 41 dB1,2
Amplitude ripple 3 dB within the design bandwidth1
Group delay ripple in the spectrum occupied by the
CMTS
75 ns within the design bandwidth1
Micro-reflections bound for dominant echo -10 dBc @ 0.5 s
-15 dBc @ 1.0 s
-20 dBc @ 1.5 s
-30 dBc @ > 1.5 s1
Carrier hum modulation Not greater than -26 dBc (5%)1
Burst noise Not longer than 25 s at a 10 Hz average rate1
11/17/11 CableLabs
21
Parameter Value
Maximum analog video carrier level at the CM input 17 dBmV
Maximum number of analog carriers 121
Table Notes:
Note 1 Measurement methods defined in [NCTA] or [CableLabs1].
Note 2 Measured relative to a QAM signal that is equal to the nominal video level in the plant.

5.2.2 Transmission Upstream
The RF channel transmission characteristics of the cable network in the upstream direction are described in Table 5
2. No combination of the following parameters will exceed any stated interface limit defined elsewhere in this
specification. Transmission is from the CM output at the customer location to the head-end.
Table 52 - Assumed Upstream RF Channel Transmission Characteristics
Parameter Value
Frequency range 5 to 42 MHz edge to edge or
5 to 85 MHz edge to edge
Carrier-to-interference plus ingress (the sum of noise,
distortion, common-path distortion and cross modulation
and the sum of discrete and broadband ingress signals,
impulse noise excluded) ratio
Not less than 25 dB1
Carrier hum modulation Not greater than -23 dBc (7.0%)
Burst noise Not longer than 10 s at a 1 KHz average rate for most
cases2,3
Amplitude ripple across upstream operating frequency
range
0.5 dB/MHz
Group delay ripple across upstream operating frequency
range
200 ns/MHz
Micro-reflections single echo -10 dBc @ 0.5 s
-20 dBc @ 1.0 s
-30 dBc @ > 1.0 s
Seasonal and diurnal reverse gain (loss) variation Not greater than 14 dB min to max
Table Notes:
Note 1 Ingress avoidance or tolerance techniques may be used to ensure operation in the presence of time-
varying discrete ingress signals that could be as high as 10 dBc. The ratios are guaranteed only within the digital
carrier channels.
Note 2 Amplitude and frequency characteristics sufficiently strong to partially or wholly mask the data carrier.
Note 3 Impulse noise levels more prevalent at lower frequencies (<15 MHz).

5.2.2.1 Availability
Typical cable network availability is considerably greater than 99%.
22 CableLabs
11/17/11
5.3 Transmission Levels
The nominal power level of the upstream CM signal(s) will be as low as possible to achieve the required margin
above noise and interference. Uniform power loading per unit bandwidth is commonly followed in setting upstream
signal levels, with specific levels established by the cable network operator to achieve the required carrier-to-noise
and carrier-to-interference ratios.
5.4 Frequency Inversion
There will be no frequency inversion in the transmission path in either the downstream or the upstream directions,
i.e., a positive change in frequency at the input to the cable network will result in a positive change in frequency at
the output.
11/17/11 CableLabs
23
6 PHYSICAL MEDIA DEPENDENT SUBLAYER SPECIFICATION
6.1 Scope
This section applies to the first technology option referred to in Section 1.1. For the second option, refer to Annex B.
This specification defines the electrical characteristics and signal processing operations for a cable modem (CM) and
Cable Modem Termination System (CMTS). It is the intent of this specification to define an interoperable CM and
CMTS such that any implementation of a CM can work with any CMTS. It is not the intent of this specification to
imply any specific implementation.
6.2 Upstream
6.2.1 Overview
The upstream Physical Media Dependent (PMD) sublayer uses an FDMA/TDMA (herein called TDMA mode) or
FDMA/TDMA/S-CDMA (herein called S-CDMA mode) burst type format, which provides six modulation rates and
multiple modulation formats. The use of TDMA or S-CDMA mode is configured by the CMTS via MAC messaging.
FDMA (frequency division multiple access) indicates that multiple RF channels are assigned in the upstream band. A
CM transmits on one or more RF channels and may be reconfigured to change channels.
A CM MUST support at least four active upstream channels (which are referred to as the Transmit Channel Set for
that CM).
The CM reports its maximum number of upstream channels capability, and certain other of its capability
characteristics, to the CMTS (Section 6.2.25).
The CM MUST be able to operate each channel in the Transmit Channel Set, simultaneously, anywhere in the
upstream band, subject to restrictions on transmit power across the channels and on reconfiguration of certain
transmit properties (see Sections 6.2.19 and 6.2.20, and subsections of each). The CMTS MUST be capable of
assigning and receiving each RF channel anywhere in the upstream band. The CMTS MUST set the number of
assigned channels and assigned channels' center frequency, and all other channel attributes. The CMTS MAY change
the number of assigned channels and the channel attributes. Each RF channel has its own set of UCD parameters as
defined in [MULPI].
TDMA (time division multiple access) indicates that upstream transmissions have a burst nature. A given RF channel
is shared by multiple CMs via the dynamic assignment of time slots. S-CDMA (synchronous code division multiple
access) indicates that multiple CMs can transmit simultaneously on the same RF channel and during the same TDMA
time slot, while being separated by different orthogonal codes.
In this document, the following naming conventions are used. For TDMA, the term "modulation rate" refers to the
RF channel symbol rate (160 to 5120 ksym/s). For S-CDMA, the term "modulation rate" refers to the "chip rate,"
which is the rate (1280 to 5120 kHz) of the individual elements (chips) of the S-CDMA spreading code. Modulation
rates are represented in units of "Hz" denoting the number of symbols per second in TDMA mode or the number of
chips per second in S-CDMA mode. The "modulation interval" is the symbol period (TDMA mode) or chip period
(S-CDMA mode) and is the reciprocal of the modulation rate. At the output of the spreader, a group of 128 chips
which comprise a single S-CDMA spreading code, and are the result of spreading a single information (QAM
constellation) symbol is referred to as a "spread symbol." The period of a spread symbol (128 chips) is called a
"spreading interval." A "burst" is a physical RF transmission that contains a single preamble plus data, and (in the
absence of preceding and following bursts) exhibits RF energy ramp-up and ramp-down.
24 CableLabs
11/17/11
In some cases logical zeros or logical ones are used to pad data blocks; this indicates data with zero-valued or one-
valued binary bits, which result in non-zero transmitted RF energy. In other cases a numerical zero is used; this
denotes, for example, symbols which result in zero transmitted RF energy (after ramp-up and ramp-down are taken
into account).
The modulation format includes pulse shaping for spectral efficiency, is carrier-frequency agile, and has selectable
output power level.
Each burst supports a flexible modulation order, modulation rate, preamble, randomization of the payload, and
programmable FEC encoding.
All of the upstream transmission parameters associated with burst transmission outputs from the CM are configurable
by the CMTS via MAC messaging. Many of the parameters are programmable on a burst-by-burst basis.
The PMD sublayer can support a near-continuous mode of transmission, wherein ramp-down of one burst MAY
overlap the ramp-up of the following burst, so that the transmitted envelope is never zero. In TDMA mode, the
system timing of the TDMA transmissions from the various CMs MUST provide that the center of the last symbol of
one burst and the center of the first symbol of the preamble of an immediately following burst are separated by at
least the duration of five symbols. The guard band MUST be greater than or equal to the duration of five symbols
plus the maximum timing error. Timing error is contributed by both the CM and CMTS. CM timing performance is
specified in Section 6.2.20.1. Maximum timing error and guard band may vary with CMTSs from different vendors.
The term guard time is similar to the guard band, except that it is measured from the end of the last symbol of one
burst to the beginning of the first symbol of the preamble of an immediately following burst. Thus, the guard time is
equal to the guard band 1.
The PMD sublayer also supports a synchronous mode of transmission when using S-CDMA, wherein ramp-down of
one burst MAY completely overlap the ramp-up of the following burst, so that the transmitted envelope is never
zero. There is no guard time for transmission on S-CDMA channels. The system timing of the S-CDMA
transmissions from the various CMs MUST provide adequate timing accuracy so that different CMs do not
appreciably interfere with each other. S-CDMA utilizes precise synchronization so that multiple CMs can transmit
simultaneously.
The upstream modulator is part of the cable modem which interfaces with the cable network. The modulator contains
the electrical-level modulation function and the digital signal-processing function; the latter provides the FEC,
preamble prepend, symbol mapping, and other processing steps.
At the Demodulator, similar to the Modulator, there are two basic functional components: the demodulation function
and the signal processing function. The Demodulator resides in the CMTS and there is one demodulation function
(not necessarily an actual physical demodulator) for each carrier frequency in use. The demodulation function
receives all bursts on a given frequency.
The demodulation function of the Demodulator accepts a varying-level signal centered around a commanded power
level and performs symbol timing and carrier recovery and tracking, burst acquisition, and demodulation.
Additionally, the demodulation function provides an estimate of burst timing relative to a reference edge, an estimate
of received signal power, may provide an estimate of signal-to-noise ratio, and may engage adaptive equalization to
mitigate the effects of a) echoes in the cable plant, b) narrowband ingress, and c) group delay. The signal-processing
function of the Demodulator performs the inverse processing of the signal-processing function of the Modulator.
This includes accepting the demodulated burst data stream and decoding, etc. The signal-processing function also
provides the edge-timing reference and gating-enable signal to the demodulators to activate the burst acquisition for
each assigned burst slot. The signal-processing function may also provide an indication of successful decoding,
decoding error, or fail-to-decode for each codeword and the number of corrected Reed-Solomon symbols in each
codeword. For every upstream burst, the CMTS has a prior knowledge of the exact burst length in modulation
intervals (see Sections 6.2.5.1, 6.2.5.2, 6.2.6, 6.2.20, and [MULPI], in the MAC Service ID section).
11/17/11 CableLabs
25
6.2.2 Signal Processing Requirements
The signal processing order for each burst packet type MUST be compatible with the sequence shown in Figure 61.
For TDMA mode, the signal processing order for each burst packet type MUST follow the order of steps in Figure
62. For S-CDMA mode, the signal processing order for each burst packet type MUST follow the order of steps in
Figure 63.
The blocks used only in S-CDMA consist of a TCM encoder, S-CDMA framer, and S-CDMA spreader. The TCM
encoder provides trellis modulation encoding of data symbols and is described in Section 6.2.9. The S-CDMA
framer maps mini-slots into code resources, provides interleaving of data symbols, and is described in Section
6.2.11.2. The S-CDMA spreader spreads S-CDMA framed symbols for transmission and is described in Section
6.2.15, "S-CDMA Spreader."

R-S
Encoder
Byte
Interleaver
(TDMA only)

Scrambler
TCM Encoder
(S-CDMA only)

Preamble
Prepend
Symbol
Mapper
Framer

Spreader Transmit
Equalizer
Filter Modulator
Burst
data in

RF out

(S-CDMA only)
(S-CDMA only)

Figure 61 - Upstream Signal-Processing Sequence

26 CableLabs
11/17/11

Packet Stream Input

Block the Data Separate Packet into Information Blocks (=data bytes in one codeword)

R-S Encode R-S (Reed-Solomon) Encode each Information Block, using shortened
codeword for last block if needed. R-S FEC can be turned off.

Byte Interleave R-S Byte Interleave. R-S Byte interleaver can be turned off.

Scramble Scramble (see Figure 68)

Preamble Prepend Prepend Preamble Symbols

Symbol Map Map the Data Stream into Modulator Symbols

Transmit Equalize Pre-Equalize the Symbol Stream

Filter Filter Symbol Stream for Spectral Shaping

Modulate Modulate at Precise Times (QPSK; 8-QAM; 16-QAM; 32-QAM; 64-QAM)

Output RF Waveform Bursts

Figure 62 - TDMA Upstream Transmission Processing

11/17/11 CableLabs
27

Packet Stream Input

Block the Data Separate Packet into Information Blocks (=data bytes in one codeword)

R-S Encode R-S (Reed-Solomon) Encode each Information Block, using shortened
codeword for last block if needed. R-S FEC can be turned off.

Scramble Scramble (see Figure 68)

TCM Encode TCM (Trellis Coded Modulation) Encode the bytes. TCM can be turned off.

Preamble Prepend Prepend Preamble Symbols

S-CDMA Framer Frame and Interleave the Data into Mini-slots.

Symbol Map Map the Data Stream into Modulator Symbols

S-CDMA Spreader Spread the Symbols. For spreader-off bursts on S-CDMA channels, spreader
can be turned off.

Transmit Equalize Pre-Equalize the Signal Stream

Filter Filter Signal for Spectral Shaping

Modulate Modulate at Precise Times (QPSK; 8-QAM; 16-QAM; 32-QAM; 64-QAM;
128-QAM/TCM only).

Output RF Waveform Bursts
Figure 63 - S-CDMA Upstream Transmission Processing

6.2.3 Modulation Formats
The upstream modulator MUST provide QPSK and 16-QAM differential encoded modulations for TDMA.
The upstream modulator MUST provide QPSK, 8-QAM, 16-QAM, 32-QAM, and 64-QAM modulations for TDMA
and S-CDMA channels.
The upstream modulator MUST provide QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, and 128-QAM TCM
encoded modulations for S-CDMA channels.
The upstream demodulator MAY support QPSK, and 16-QAM differential modulation for TDMA.
The upstream demodulator MUST support QPSK, 16-QAM, and 64-QAM modulations for TDMA and S-CDMA
channels.
28 CableLabs
11/17/11
The upstream demodulator MAY support 8-QAM and 32-QAM modulation for TDMA and S-CDMA channels.
The upstream demodulator MAY support QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, and 128-QAM TCM
encoded modulations for S-CDMA channels.
6.2.4 R-S Encode
6.2.4.1 R-S Encode Modes
The upstream modulator MUST be able to provide the following selections: Reed-Solomon codes over GF(256) with
T = 1 to 16 or no R-S coding.
The following Reed-Solomon generator polynomial MUST be supported:
g(x) = (x+
0
) (x+
1
)...(x+
2T-1
)
where the primitive element alpha is 0x02 hex.
The following Reed-Solomon primitive polynomial MUST be supported:
p(x) = x
8
+ x
4
+ x
3
+ x
2
+ 1
The upstream modulator MUST provide code words from a minimum size of 18 bytes (16 information bytes [k] plus
two parity bytes for T = 1 error correction) to a maximum size of 255 bytes (k-bytes plus parity-bytes). The
minimum uncoded word size MUST be one byte.
In Shortened Last Codeword mode, the CM MUST provide the last codeword of a burst shortened from the assigned
length of k data bytes per codeword as described in Sections 6.2.5.1.3 and 6.2.6.
The value of T MUST be configured in response to the Upstream Channel Descriptor from the CMTS.
6.2.4.2 R-S Bit-to-Symbol Ordering
The input to the Reed-Solomon Encoder is logically a serial bit stream from the MAC layer of the CM, and the first
bit of the stream MUST be mapped into the MSB of the first Reed-Solomon symbol into the encoder. The MSB of
the first symbol out of the encoder MUST be mapped into the first bit of the serial bit stream fed to the Scrambler.
Note: The MAC byte-to-serial upstream convention calls for the byte LSB to be mapped into the first bit
of the serial bit stream.
6.2.5 Upstream R-S Frame Structure for DOCSIS 3.0 Multiple Transmit Channel mode Enabled
This section applies to CMs operating in DOCSIS 3.0 Multiple Transmit Channel mode enabled in the upstream
direction.
Figure 64 shows two examples of the R-S frame structure: one where the packet length equals the number of
information bytes in a codeword, and another where the packet length is longer than the number of information bytes
in one codeword, but less than in two code words. Example 1 illustrates the fixed codeword-length mode, and
Example 2 illustrates the shortened last codeword mode. These modes are defined in Section 6.2.5.1.
11/17/11 CableLabs
29
Preamble
Preamble Two Codewords
Guard
Time
Guard
Time
FEC
Parity
FEC
Parity
Packet Data
Empty up to
Next Mini-Slot
Boundary
First k Bytes
of Packet
Last k Bytes
of Packet
k-k bytes of
one-fill
FEC
Parity
Example 1. Packet length = number of information bytes in codeword = k
Example 2. Packet length = k + remaining information bytes in 2
nd
codeword = k + k k + k 2k
One Codeword
mini-slot
boundary

Figure 64 - Example Frame Structures with Flexible Burst Length Mode, DOCSIS 3.0 Operation

6.2.5.1 R-S Codeword Length
When R-S FEC is enabled, the CM operates in either fix-length codeword mode or in shortened-last codeword mode.
The minimum number of information bytes in a codeword in either mode is 16. Shortened-last codeword mode only
provides a benefit when the number of bytes in a codeword is greater than the minimum of 16 bytes.
The intent of the following sections is to define rules and conventions such that the CMTS PHY knows what to
expect regarding the R-S FEC framing in both fixed codeword length and shortened last codeword modes. Shortened
last codeword mode MUST NOT be used for Initial Maintenance (broadcast or unicast).
6.2.5.1.1 Burst Size
For an allocation of mini-slots (in both contention and non-contention regions), the requirements of Sections
6.2.5.1.2 and 6.2.5.1.3 apply to a burst transmitted in that allocation. Regardless of the size of the allocation, the size
of the burst MUST be as specified in Table 61 below.
Table 61 - Burst Size
IUC Burst Size
1, 3 Minimum number of mini-slots required for message transmission including burst overhead. Burst
overhead includes pre-amble, R-S parity bytes, TCM return-to-zero bits, and guard time if
applicable.
2 Number of mini-slots specified in the Well-Known Multicast SID.
4-6, 9-11 Number of mini-slots allocated.

30 CableLabs
11/17/11
6.2.5.1.2 Fixed Codeword Length
With the fixed-length code words, after all the data are encoded, one-fill MUST occur in this codeword if necessary
to reach the assigned k data bytes per codeword. Additionally, one-fill MUST continue up to the point when no
additional fixed-length code words can be inserted before the end of the burst specified in Table 61 above,
accounting for preamble, FEC parity, return-to-zero bits, and guard-time symbols (if any).
6.2.5.1.3 Shortened Last Codeword
As shown in Figure 64, let k' = the number of information bytes that remain after partitioning the information bytes
of the burst into full-length (k burst data bytes) code words. The value of k' is less than k. Given operation in a
shortened last codeword mode, let k" = the number of burst data bytes plus one-fill bytes in the shortened last
codeword. In shortened codeword mode, the CM MUST encode the data bytes of the burst (including MAC Header)
using the assigned codeword size (k information bytes per codeword) until 1) all the data are encoded, or 2) a
remainder of data bytes is left over which is less than k. Shortened last code words MUST NOT have less than 16
information bytes, and this is to be considered when CMs make requests of mini-slots. In shortened last codeword
mode, the CM MUST one-fill data if necessary up to the Burst Size specified in Table 61 above accounting for
preamble, FEC parity, return-to-zero bits and guard-time symbols (if any). Therefore, in many cases, only k" - k' one-
fill bytes are necessary with 16 k" k and k' k".
Generally, the CM MUST one-fill data until the point when no additional fixed-length code words can be inserted
before the end of the burst specified in Table 61 above, accounting for preamble, FEC parity, return-to-zero bits,
and guard-time symbols (if any). Then, if possible, a shortened last codeword of one-fill MUST be inserted to fit into
the last mini-slot.
If, after one-fill of additional code words with k information bytes, there are less than 16 bytes remaining before the
end of the burst specified in Table 61 above, accounting for preamble, FEC parity, return-to-zero bits and guard-
time symbols (if any), then the CM MUST NOT create this last shortened codeword.
6.2.5.2 R-S FEC Disabled
When T = 0 (no FEC parity bytes), the CM MUST one-fill in full bytes to the end of the burst specified in Section
6.2.5.1.1 above, accounting for preamble, return-to-zero bits, and guard-time symbols (if any).
6.2.6 Upstream R-S Frame Structure for DOCSIS 3.0 Multiple Transmit Channel mode Not
Enabled
When Multiple Transmit Channel mode is not enabled, all requirements in Section 6.2.5 for R-S frame structure in
Multiple Transmit Channel Operation apply, except that zero-fill MUST be used instead of one-fill in Sections
6.2.5.1.2, 6.2.5.1.3, 6.2.5.2, and in Figure 64.
6.2.7 TDMA Byte Interleaver
R-S codeword interleaving in a byte (R-S symbol) format MUST be performed after R-S encoding on a TDMA
channel. The byte interleaver changes the order of the bytes at the R-S encoder output, i.e., it performs an operation
of byte permutation. At the receiver side, the original order of bytes is restored prior to the R-S decoding. Therefore,
if some consecutive bytes were corrupted by burst noise, they are spread between various R-S code words, averaging
the number of erroneous bytes in each codeword. The interleaver is a block interleaver type, i.e., the permutation is
achieved by filling a table row-wise (one row per R-S codeword), and reading it column-wise. The total memory size
allocated for the table is 2048 bytes.
The byte interleaver is disabled when the R-S encoder is turned off (T = 0).
11/17/11 CableLabs
31
6.2.7.1 Byte Interleaver Parameters
The interleaver operating parameters described in Table 62 determine the operation of the interleaver for every
burst.
Table 62 - Interleaver Operating Parameters
Parameter Definition Allowed Values
N
r
Interleaver Width (R-S Codeword Length, k+2*T) 18 to 255
I
r
Interleaver Depth 0 - Dynamic Mode
1 - No Interleaving
2 to floor(2048/N
r
) - Fixed Mode
B
r
Interleaver Block Size 2*N
r
to 2048
N
f
Packet Size (in bytes, including FEC) 18 bytes

The CMTS and CM MUST use the interleaver parameters within the allowed values in Table 62 with the following
additional restrictions:
1. N
r
and I
r
MUST be chosen such that N
r
I
r
2048 (in other words, for a given N
r
, the maximal value of I
r
is
I
r
,max=floor(2048/N
r
).
2. N
r
MUST be identical to the R-S codeword length (i.e., k+2T).
3. B
r
is effective only when I
r
=0. This mode is called dynamic mode.
4. When I
r
=1, interleaving is disabled.
N
r
, I
r
, and B
r
are specified in the burst profile, and N
f
is implied in the MAP message.
6.2.7.2 Interleaver Operating Modes
The interleaver MUST support both an operating mode in which the block size is fixed, as well as a dynamic mode
in which the interleaver depth is determined based on the burst size.
6.2.7.2.1 Fixed Mode
The R-S encoded data bytes of the packet are first divided into interleaver blocks of N
r
I
r
bytes (i.e., blocks of I
r
R-S
code words each). The size of the last interleaver block may be smaller when the packet length is not an integer
multiple of N
r
I
r
. Each interleaver block is interleaved separately.
Each interleaver block is filled into a table with I
r
rows and N
r
columns. The data is written row by row (from left to
right). Therefore, each row corresponds to one R-S codeword. The bytes are read column by column (from top to
bottom). The interleaver operation is demonstrated in Figure 65.
32 CableLabs
11/17/11
Write
R
e
a
d
C
1
(1) C
1
(2)
C
2
(1) C
2
(2)
C
Ir
(1) C
Ir
(2)
Input sequence: C
1
(1),...C
1
(N
r
),C
2
(1),....C
2
(N
r
),C
3
(1).....C
Ir
(N
r
)
Output sequence: C
1
(1),C
2
(1)...C
Ir
(1),C
1
(2),....C
Ir
(2),C
1
(3).....C
Ir
(N
r
)
C
1
(N
r
)
C
2
(N
r
)
C
Ir
(N
r
)

Figure 65 - Byte Interleaver Operation

The last interleaver block might have fewer rows than I
r
. If the shortened last codeword mode is applied, then the last
row might have fewer elements than N
r
. In these cases, the interleaver table is read column by column, skipping the
empty elements of the table. The interleaver operation for the last interleaver block is demonstrated in Figure 66.
Write
R
e
a
d
Inputsequence:C
1
(1),...C
1
(N
r
),C
2
(1),....C
2
(N
r
),C
3
(1).....C
I'
(1)......C
I'
(N')
Outputsequence:
C
1
(1),C
2
(1)...C
I'
(1),C
1
(2)...C
I'
(2)...C
1
(N')...C
I'
(N'),C
1
(N'+1)....C
I'-1
(N'+1),C
1
(N'+2)...C
i'-1
(N'+2)...C
1
(N
r
)...C
I'-1
(N
r
)
C
1
(1) C
1
(2) C
1
(N
r
)
C
2
(1) C
2
(2) C
2
(N
r
)
C
I'
(1) C
I'
(N') C
I'
(2)
C
1
(N')
C
i'-1
(N
r
) C
i'-1
(N')

Figure 66 - Interleaver Operation for Last Interleaver Block (with Shortened Last Codeword)

6.2.7.2.2 Dynamic Mode
In the fixed mode, the interleaving depth of the last interleaving block of a packet (I' in Figure 66) may be as small
as one, resulting in low burst noise robustness for this block. In the dynamic mode, the depths of the interleaver
blocks are chosen such that all blocks have approximately the same depth to achieve nearly optimal burst noise
robustness (for the given block size).
The R-S encoded data bytes of the packet are first divided into
0
s
N interleaver blocks. The size of the i
th
interleaver
block is N
r
*I
r
(i)
bytes (i.e., a block of I
r
(i)
R-S code words). The size of the last interleaver block may be smaller in the
11/17/11 CableLabs
33
shortened last codeword mode. Each interleaver block is interleaved separately (see the equations for
0
s
N and I
r
(i)
in
Section 6.2.7.2.2.1).
The i
th
interleaver block is filled into a table with I
r
(i)
rows and N
r
columns. The data is written row-wise (from left to
right). Therefore, each row corresponds to one R-S codeword. The bytes are read column-wise (from top to bottom).
The interleaver operation is demonstrated in Figure 65 (except that there are I
r
(i)
rows instead of I
r
).
If the shortened last codeword mode is applied, then the last row might have fewer elements than N
r
. In this case, the
interleaver table is read column by column, skipping the empty elements of the table. The interleaver operation for
the last interleaver block is demonstrated in Figure 66 (except that there are
( )
0
s
N
r
I rows instead of I').
6.2.7.2.2.1 Dynamic Mode Calculations
Dynamic mode calculations are illustrated below.
0
s
N and
) (i
r
I are determined by the following equations:
Total number of interleaver rows: ) / (
0
r f tot
N N ceil I = .
Maximal number of rows per segment: ) / (
max , r r r
N B floor I = .
Number of segments: ) / (
max ,
0 0
r tot s
I I ceil N =
Interleaver depth of first block: ) / (
0 0 1
s tot r
N I floor I =
No. of blocks with depth of
1
r
I :
0 1 0
) 1 (
tot r s
I I N M + =
Then for segment i,
) (i
r
I is calculated as follows (
0
s
N ... 1 i = ):

+ = +
=
=
0 1
1
) (
,..., 1 , 1
,..., 1 ,
s r
r i
r
N M i I
M i I
I

Figure 67 - T Mode Calculations

6.2.8 Scrambler (Randomizer)
The upstream modulator MUST implement a scrambler (shown in Figure 68) where the 15-bit seed value is
arbitrarily programmable.
6

At the beginning of each burst, the register is cleared and the seed value is loaded. The seed value MUST be used to
calculate the scrambler bit which is combined in an XOR with the first bit of data of each burst (which is the MSB of
the first symbol following the last symbol of the preamble).
The scrambler seed value MUST be configured in response to the Upstream Channel Descriptor from the CMTS.

6
Revised text per PHYv3.0-N-06.0341-1 by GO on 1/12/07.
34 CableLabs
11/17/11
The polynomial MUST be x
15
+ x
14
+ 1.
SEED
LSB
SEED
MSB
SEEDLOAD/FRAME RESET
DELAY
ELEMENT
1
DELAY
ELEMENT
2
DELAY
ELEMENT
3
DELAY
ELEMENT
N-2
DELAY
ELEMENT
N-1
DELAY
ELEMENT
N
RANDOMIZER
DATA INPUT
N=15
XOR
RANDOMIZER
DATA OUTPUT
XOR

Figure 68 - Scrambler Structure

6.2.9 TCM Encoder
R-S symbol interleaving is commonly included between the TCM and R-S blocks to preserve coding gain in the
presence of bursts of errors produced at the output of the TCM decoder. This interleaver was not included in the
original baseline S-CDMA proposal to reduce memory requirements at the expense of coding gain.
In S-CDMA mode, the CM MUST support trellis coded modulation for transmission of m = 1, 2, 3, 4, 5, and 6 bits
per symbol with QPSK0, 8-QAM, 16-QAM, 32-QAM, 64-QAM, and 128-QAM constellations, respectively.
Support of TCM in the CMTS is optional.
Figure 69 shows the employed 8-state TCM encoder. The encoding operation causes a mapping of m input bits into
m+1 output bits for input into the symbol mapping block. The systematic convolutional encoder adds the coded bit x
1

= s
0
to the input bits i
m
,i
3
,i
2
,i
1
. For m = 1, only input bit i
1
is used (i
2
= 0), and encoding is reduced to rate-1/2 coding.
D D D
x
m+1
:
x
4
x
3
x
2
i
m
:
i
3
i
2
i
1
s
2
s
1
s
0
=x
1
S-CDMA
Framing
symbol
labels

Figure 69 - Convolutional encoder
The initial state of the TCM encoder MUST be the zero state. The zero state MUST be reached again with the last
encoded symbol.
11/17/11 CableLabs
35
To return to the zero state from all possible Trellis paths, if m = 1 (QPSK) three tail symbols (n
t
= 3) MUST be
generated with input bit i
1
set to i
1
= s
1
. By inspection of Figure 69, after three symbols the state bits s
2
, s
1
, and s
0
=
x
1
will be zero. Tail symbols are extra symbols, which carry no information.
If m = 2, to return to the zero state from all possible Trellis paths, two tail symbols (n
t
=2) MUST be generated. The
input bits i
2
i
1
MUST be set such that the zero state is reached after two symbols. If the first symbol is set to i
2
= 0, i
1
=
s
1
, and the second (final) symbol to i
2
= s
2
, i
1
= s
1
after these two symbols the state bits s
2
, s
1
, and s
0
= x
1
will be zero.
If m 3, the uncoded bits i
m
,..,i
3
MUST be used for information encoding, when this is possible. Otherwise, uncoded
bits MUST be set to zero. The number of tail symbols carrying no information depends on the ending conditions and
can vary between zero and two (0 n
t
2).
6.2.9.1 Byte to TCM Symbol Mapping
The mapping of bytes to TCM symbols is done such that each byte is mapped entirely to the uncoded bits
i
m
,...,i
3
, or entirely to the convolutional encoder input bits i
2
i
1
. The decision is made sequentially for each byte
using the rule that the byte assignment should lead to the shortest packet of symbols including tail symbols, if the
current byte were the last byte to be encoded. This rule results in the repetitive patterns of byte assignments to label
bits shown in Figure 610 for m=1 to 6. In the figure bit i
m
is at the top and bit i
1
is at the bottom.

The MSB (i
m
) MUST be the first bit in the serial data fed into the uncoded input bits (i.e., i
m
to i
3
). The MSB (i
2
)
MUST be the first bit in the serial data fed into the coded input bits.
Figure 611 illustrates the byte assignments for Trellis-coded 64-QAM modulation by two examples. Notice that
bytes are assigned in a repetitive pattern of five bytes. In the first example, N
f
is divisible by five. In this case two tail
symbols are appended. In the second example, N
f
is not divisible by five and no tail symbols are required. The bits
needed for returning to the zero state are available in symbols still carrying information.
m=6(128QAM)
m=5(64QAM)
m=4(32QAM)
m=3(16QAM)
m=2(8QAM)
m=1(QPSK)
1
1 2
1 3
2
1
2
3 5
1
2
1 2 4
2
3
1

Figure 610 - Repetitive Patterns of Byte Mapping to Symbol Map Bits for TCM

36 CableLabs
11/17/11
0
zero fill
bits
bits to return
to zero state
redundant
code bits
x
6
x
5
x
4
x
3
x
2
x
1
n
p
preamble
symbols
n
i
TCM
symbols
msb
1
msb
3
msb
2
msb
4
msb
8
msb
6
msb
N
f
-2
0
0
0
0
msb
N
f
msb
N
f
-1
msb
5
n
t
= 2
tail symbols
0
0
msb
1
msb
3
msb
2
msb
4
msb
8
msb
6
msb
N
f
-2
0
msb
N
f
msb
N
f
-1
msb
5
zero
initial
state
zero
final
state
0
0
0
.........
.........
.........
.........
.........
.........
.........
.........
...............
..
..
..
n
i
TCM symbols, no
tail symbols (n
t
= 0)

Figure 611 - Example Byte to Bit Assignment for 64-QAM

The CM MUST place the return-to-zero bits right after the last TCM-coded data sub-symbol, that is, the last coded
sub-symbol corresponding to the parity bytes of the last shortened or fixed R-S codeword including any zero-filled
(for operation not in MTC Mode) or one-filled (for operation in MTC Mode) R-S code words in the grant. The rest
of the TCM-coded bits MUST be filled with zeros.

Figure 612 illustrates the placing of return-to-zero bits for 64-QAM when the last transmitted byte is #1. The first
two pairs of x
2
and x
3
are the return to zero bits, and the last empty coded pair is zero-filled.
7

7
Revised paragraph per PHYv3.0-N-08.0701-2 on 12/23/08 by CP
11/17/11 CableLabs
37
Uncoded bits
Zero fill bits
Bits to return to zero state
u
0
r
u u u
u u
u
u
u
0
0 0
0 r
r
r
i
5
x
6
i
4
i
3
i
2
i
1
x
5
x
4
x
3
x
2
x
1

Figure 612 - Example of return to zero bits followed by "0"

6.2.10 Preamble Prepend
The upstream PMD sublayer MUST support a variable-length preamble field that is prepended to the data after it has
been randomized, Reed-Solomon encoded, and TCM encoded.
The first bit of the Preamble Pattern is the first bit into the symbol mapper (see Section 6.2.14). The first bit of the
Preamble Pattern is designated by the Preamble Value Offset. The preamble is interleaved by the framer in S-CDMA
mode.
The preamble sequence MUST be programmable. For DOCSIS 2.0/3.0 bursts (bursts encoded using a Type 5 burst
descriptor), the preamble MUST use the QPSK0 or QPSK1 constellation (per Figure 620 and Figure 621) with
preamble length 0, 2, 4, 6,..., or 1536 bits (maximum 768 QPSK symbols). For DOCSIS 1.x compatible bursts (Type
4 burst descriptor) that use QPSK modulation, the preamble and data MUST use the QPSK0 constellation with
preamble length 0, 2, 4, 6,..., or 1024 bits (maximum 512 QPSK symbols). For DOCSIS 1.x compatible bursts (Type
4 burst descriptor) that use 16-QAM modulation, the preamble and data MUST use the 16-QAM constellation with
preamble length 0, 4, 8, 12,..., or 1024 bits (maximum 256 16-QAM symbols).
The preamble length and value MUST be configured in response to the Upstream Channel Descriptor message
transmitted by the CMTS.
38 CableLabs
11/17/11
6.2.11 Modulation Rates
6.2.11.1 DOCSIS 3.0 Modulation Rates
In TDMA and S-CDMA modes, the CM upstream modulator MUST provide all modulations at 1280, 2560, and
5120 kHz.
In TDMA and S-CDMA modes, the CMTS upstream demodulator MUST be able to support demodulation at 1280,
2560, and 5120 kHz.
This variety of modulation rates, and flexibility in setting upstream carrier frequencies, permits operators to position
carriers in gaps in the pattern of narrowband ingress.
The modulation rate for each upstream channel is defined in an Upstream Channel Descriptor (UCD) MAC message.
All CMs using that upstream channel MUST use the defined modulation rate for upstream transmissions.
6.2.11.2 Backward Compatibility Modulation Rates
Operating with a DOCSIS 1.x or DOCSIS 2.0 CMTS, or with a CMTS operating in one of these modes, in TDMA
and S-CDMA modes, the CM upstream modulator MUST provide all modulations at 1280, 2560, and 5120 kHz.
In addition, with such pre-3.0-DOCSIS CMTS operation, for TDMA mode, the CM upstream modulator MAY
provide all modulations at 160, 320, and 640 kHz.
In pre-3.0-DOCSIS CMTS operation, for both TDMA and S-CDMA modes, the CMTS upstream demodulator
MUST be able to support demodulation at 1280, 2560, and 5120 kHz. In pre-3.0-DOCSIS CMTS operation, for
TDMA mode, the CMTS upstream demodulator MAY support demodulation at 160, 320, and 640 kHz.
6.2.12 S-CDMA Framer and Interleaver
6.2.12.1 S-CDMA Framing Considerations
The S-CDMA mode of the PHY layer accepts data presented to it for transmission from the MAC layer. This data is
presented as bursts of n mini-slots. These bursts are mapped within the PHY layer to a combination of spreading
codes and time slots, in order to exploit the multi-dimensional spreading of information by the S-CDMA mode.
There are various adjustable parameters in the upstream channel parameters and upstream burst attributes that allow
controlling the mini-slot to physical layer mapping, as well as tuning the channel to accommodate a variety of
channel conditions, noise characteristics, capacities, reliability levels, and latency requirements.
When operating in S-CDMA mode, data is transmitted in two dimensions: codes and time. For this reason, data to be
transmitted is grouped into two-dimensional rectangular frames prior to transmission.
At the physical layer, data is sent over an array of up to 128 spreading codes. There is a programmable number of
spreading intervals per frame, as shown in Figure 613 below. A spreading interval is the time required to transmit
one symbol per code across all 128 codes in S-CDMA mode. Note that the specific codes which are used and the
details of the spreading operation are described in detail in Section 6.2.15.
A burst from a particular CM may be transmitted on two or more codes in one or more frames. A frame may contain
bursts transmitted simultaneously from multiple CMs (each on a separate subset of the codes) as defined by the MAP
message.
11/17/11 CableLabs
39
6.2.12.2 Mini-slot Numbering
In normal operation, the MAC will request the PHY to transmit a burst of length n mini-slots, starting at mini-slot m,
as defined by the MAP. All CMs and the CMTS MUST have a common protocol of how mini-slots are numbered,
and how they are mapped onto the physical layer framing structure. This common protocol is obtained from
information in the SYNC and Upstream Channel Descriptor (UCD) messages.
Mini-slots are mapped onto frames starting at the first active code, are numbered sequentially through the remainder
of the frame, and then wrap to the next sequential frame (see Section 6.2.12.2.2 for mini-slot numbering examples).
Mini-slots are mapped onto a group of consecutive codes.
The CMTS and the CMs require a common protocol for mini-slot numbering. For operation on a TDMA channel,
this is achieved solely through recovery of the timestamp. Since the time duration of an S-CDMA frame is not
necessarily a power-of-2 multiple of the 10.24 MHz reference, the timestamp rollover (at 2
32
counts) is not
necessarily at an S-CDMA frame boundary. Therefore, an additional synchronization step is required.
The CMTS is required to identify frame boundaries relative to the timestamp counter on a periodic basis. This is
called the timestamp snapshot and has to be sent in the UCD for each upstream S-CDMA channel.
The CMTS is required to maintain a frame counter and a mini-slot counter. The CMTS has to sample these values
along with the timestamp, on a frame boundary, as shown in Figure 613 below. The CMTS is required to obtain a
new sample prior to sending each UCD message.
mini-slot m+62
mini-slot m+125 mini-slot m+188
mini-slot m mini-slot m+63 mini-slot m+126
K spreading intervals K spreading intervals K spreading intervals
code 127
code 126
code 3
code 2
code 1
code 0
t
frame f frame f+1 frame f+2
t
4 t
2
t
1 t
3
Timestamp counts
timestamp count mini-slot number frame number
32 bits 32 bits 8 bits
Timestamp snapshot
Included in
UCD

Figure 613 - Timestamp Snapshot

Each CM MUST maintain a timestamp counter, mini-slot counter, and frame counter functionally identical to the
CMTS.
40 CableLabs
11/17/11
From the UCD message, the CM receives the CMTS timestamp snapshot and parameters from which it can calculate
the number of time counts per S-CDMA frame. Using modulo arithmetic, the CM can then calculate accurate values
for timestamp, mini-slot, and frame counters at any point into the future.
The CM can then update its local mini-slot and frame counters at an appropriate timestamp counter value. At this
point, the CM representation of mini-slots and frames are aligned with those in the CMTS.
The CMTS and CM MUST implement a 32-bit timestamp counter, a 32-bit mini-slot counter, and an 8-bit frame
counter, as follows:
The mini-slot counter MUST contain the value of the first mini-slot of the frame when it is sampled. It MAY be
incremented by the number of mini-slots per frame, once per frame interval. The mini-slot counter will use all 32
bits and mini-slot numbers will, therefore, range from 0 to 2
32
-1.
The only specified function for the frame counter is to reset the code hopping sequence at the frame 0 (modulo-
256) boundary, as defined in Section 6.2.15.1.
The frame structure above relates to the entire upstream and not necessarily to the transmission from a single CM.
The codes are resources which are allocated to CMs over each S-CDMA frame. The assignment of codes to CMs is
performed by the framer as it assigns a burst of symbols a particular order in the two-dimensional matrix of codes
and time. This symbol sequencing is described in detail in Section 6.2.13.
6.2.12.2.1 Mini-slot Numbering Parameters in UCD
There are three parameters specified in the UCD that define mini-slot mapping: spreading intervals per frame, codes
per mini-slot, and number of active codes.
Spreading intervals per frame
The number of spreading intervals per frame, K, (along with the signaling rate), 1/T
s
, define the time duration of an
S-CDMA frame, T
fr
:
T
fr
= K * 128 * T
s
Note that the code length in the above equation is always 128, regardless of how many codes are currently active.
The valid range of the spreading intervals per frame parameter is 1 to 32.
Codes per mini-slot
In conjunction with the spreading intervals per frame parameter, the codes per mini-slot (C
ms
) parameter defines the
total number of symbols per mini-slot and, therefore, the mini-slot capacity. The mini-slot capacity, S
ms
, is given in
symbols by the following expression:
S
ms
= K * C
ms
The lower limit on mini-slot capacity is 16 symbols. However, the mini-slot also has to be large enough to allow the
transmission of the largest-sized data PDU (including physical layer overhead) in 255 mini-slots. The maximum
mini-slot capacity in symbols is given by the product of the maximum number of spreading intervals per frame and
the maximum number of codes per mini-slot (32x32 = 1024 symbols). The valid range of the codes per mini-slot
parameter is 2 to 32.
Selectable Active Codes (SAC)
The number of active codes parameter N
a
allows the number of codes used to carry data to be less than or equal to
128. When N
a
< 128, the active codes may be chosen using the following two modes.
11/17/11 CableLabs
41
Selectable Active Codes Mode 1: The low-numbered codes starting with code 0 are not used, as shown below in
Figure 615 (an example of 126 active codes and code hopping off).
Selectable Active Codes Mode 2: The active codes are selectable via a 128-bit string. The first element in the string
corresponds to code 0 (the all-ones code). A "1" element in the string indicates an active code, and a "0" indicates an
unused code.
There are several reasons why it may be desirable to reduce the number of active codes:
Code 0 does not have the same spreading properties as the other codes and, therefore, under certain colored
noise conditions, will degrade performance.
In extremely noisy plant conditions, a reduction in the number of active codes (along with the corresponding
increase in power per code for the remaining codes) can allow reliable operation at reduced capacities.
Reduction in active codes from 128 to 64 results in a 3 dB improvement in SNR.
The number of mini-slots per S-CDMA frame MUST be an integer. Therefore, the codes per mini-slot and
number of active codes parameters MUST be chosen to result in an integral number of mini-slots per frame.
When N
a
64, the S-CDMA frame consists of more than 1 mini-slot, since the number of codes per mini-slot is in
the range 2 to 32. This implies that N
a
is non-prime. The prime numbers between 64 and 128 are {67, 71, 73, 79, 83,
89, 97, 101, 103, 107, 109, 113, and 127}.
The CM MUST support any non-prime number of active codes in the range 64 N
a
128. The CM MUST support
Selectable Active Codes Mode 1. The CM SHOULD support Selectable Active Codes Mode 2.
The CMTS MUST support 126 and 128 active codes. The CMTS MUST support Selectable Active Codes Mode 1.
The CMTS MAY support Selectable Active Codes Mode 2.
6.2.12.2.2 Mini-slot Numbering Examples
A typical mini-slot numbering example with N
a
= 128 active codes and code hopping off is shown in Figure 614. In
this example, there are two codes per mini-slot defined. The number of codes per mini-slot is an adjustable
parameter (via the UCD) to allow flexibility in determining the effective capacity of each mini-slot.
mini-slot m+63
mini-slot m+1 mini-slot m+65 mini-slot m+129
code 127
code 126
code 3
code 2
code 1
code 0

Figure 614 - Mini-slot Mapping with Two Codes per mini-slot, 128 Active Codes

42 CableLabs
11/17/11
A second example, using three codes per mini-slot, with code hopping off, is shown in Figure 615. Since it is
required that there be an integral number of mini-slots per frame, the number of active codes N
a
has been restricted
to 126, via either Selective Active Codes Mode 1 or 2. In this example, a tradeoff has been made to increase
mapping flexibility at the expense of a small reduction in channel capacity (the ratio 2/128).
code 127
code 126
code 7
code 6
code 4
code 3
mini-slot m+41
mini-slot m+85 mini-slot m+1 mini-slot m+43
code 125
code 5
code 2
code 1
code 0

Figure 615 - Mini-slot Mapping with Three Codes per mini-slot, 126 Active Codes

Figure 616 shows an example of N
a
= 124 active codes with codes 0, 1, 5, and 125 unused; Selectable Active Codes
Mode 2 and code hopping off. The original spreading code matrix has been re-ordered such that the unused codes
are moved to the bottom rows. The active codes are arranged in the matrix in increasing order from bottom to top, as
are the unused codes. The mini-slots are defined on the active codes only, as shown in Figure 616.
mini-slotm
mini-slotm+1
mini-slotm+30
code0
code1
code5
code125
code2
code3
code4
code6
code7
code8
code9
code10
code127
code126
code124
code123
unused
codes
mini-slotm+31
mini-slotm+32
mini-slotm+61
mini-slotm+62
mini-slotm+63
mini-slotm+92
framef framef+1 framef+2
Kspreadingintervals Kspreadingintervals Kspreadingintervals
active
codes

Figure 616 - Mini-slot Mapping with Four Codes per Mini-slot, 124 Active Codes, Codes 0, 1, 5, and 125
Unused; Selectable Active Codes Mode 2

11/17/11 CableLabs
43
There is no implication that physical layer processing is performed on a per-mini-slot basis. As in a TDMA channel,
the physical layer is concerned only with the burst start time (mini-slot number) and the burst length.
6.2.12.3 Transmission Time
Ideally, all the mini-slots contained in one S-CDMA frame are received simultaneously. These mini-slots may be
transmitted from a single CM or may be transmitted from multiple CMs, as defined by the bandwidth allocation
MAP message and the mini-slot mapping configuration settings (from the UCD). Note that a single CM may have
more than one allocation active in a single S-CDMA frame.
6.2.12.4 Latency Considerations
S-CDMA frame timing is derived directly from (is phase locked to) the 10.24 MHz CMTS Master Clock. Based on
the allowable modulation rates, and the fact that there are 128 modulation intervals (chip periods) in a spreading
interval, the S-CDMA frame time MUST always be a multiple of 25 s.
The number of spreading intervals per frame and the modulation rate exactly define the S-CDMA frame duration. As
a specific example, a burst profile defined with 10 spreading intervals per frame, with a modulation rate of 2.56
MHz, would result in a frame duration of 500 s.
The amount of additional upstream latency added by the use of S-CDMA mode is approximately one S-CDMA
frame, with the exact value described in Section 6.2.18.
6.2.12.5 Spreader-off Bursts for Maintenance on S-CDMA channel
Spreader-off bursts are defined as bursts on an S-CDMA channel which has attributes that specify the spreader be
turned off. For a spreader-off burst, both the S-CDMA framer and S-CDMA spreader are bypassed. The Initial
Maintenance burst type MUST be specified (via UCD) to use spreader-off bursts. The Station Maintenance burst
type MAY be specified (via UCD) to use spreader-off or spreader-on bursts. The CM MUST support both spreader-
on and spreader-off modes for Station Maintenance bursts. All remaining IUC burst types MUST be specified (via
UCD) to use spreader-on bursts. The S-CDMA channel will be programmed (via UCD) for C
ms
codes per mini-slot,
p number of active codes, K spreading intervals per S-CDMA frame, and a resultant s mini-slots per frame, where
s=p/C
ms
.
Then each S-CDMA frame, where a transmission with the spreader off is to occur, will contain exactly s mini-slots,
where each mini-slot consists of C
ms
*K symbols, as shown in Figure 617.
In the case where the number of active codes (p) is less than 128, the frame will still contain exactly s mini-slots,
where each mini-slot consists of C
ms
*K symbols. The first mini-slot of a frame will start with the first symbol of the
frame. If a burst spans multiple frames, the burst will start relative to the first frame and continue without interruption
into the next frame.
Spreader-off bursts for Station Maintenance Regions (IUC 4) MUST be padded with zero data symbols from the end
of the R-S encoded data until the end of the burst as defined by the burst boundaries of Section 6.2.5.1.1. Spreader-
off bursts for Initial Maintenance Regions (IUC3) MUST be padded with zero data symbols from the end of the R-S
encoded data until the end of the burst as defined by the burst boundaries of Section 6.2.5.1.1. Differential encoding
and R-S byte interleaving MUST NOT be used with spreader-off bursts on S-CDMA channels.
44 CableLabs
11/17/11
m
i
n
i
-
s
l
o
t
m
+
6
4
mini-slotm+191
mini-slotm+129
mini-slotm+128
mini-slotm+63
mini-slotm+1
mini-slotm
... ... ...
code0
code1
code2
code3
code126
code127
framef framef+1 framef+2
KSpreadingIntervals 128*KModulationIntervals KSpreadingIntervals
Spreader-onframe Spreader-offframe Spreader-onframe
m
i
n
i
-
s
l
o
t
m
+
6
5
m
i
n
i
-
s
l
o
t
m
+
6
6
m
i
n
i
-
s
l
o
t
m
+
6
7
m
i
n
i
-
s
l
o
t
m
+
6
8
m
i
n
i
-
s
l
o
t
m
+
1
2
7

Figure 617 - S-CDMA Spreader-on and Spreader-off Intervals

The CMTS scheduler MUST ensure that the spreader-off interval is aligned to the start of an S-CDMA frame, occurs
completely within one or more S-CDMA frames, and MUST ensure that no spreader-on bursts are scheduled during
these same frames. The CMTS scheduler MUST grant at most one spreader-off burst per CM per frame. It is the
responsibility of the CMTS to allocate mini-slots to the NULL SID, as required to prevent interference between
bursts (i.e., before and after spreader-off bursts when the CM might not be sufficiently synchronized). Specifically,
the CMTS MUST issue a NULL grant (to the NULL SID) of 1 mini-slot immediately before each spreader-off burst,
which corresponds to either Station Maintenance, or Unicast Initial Maintenance. The CMTS MUST also issue a
NULL grant (to the NULL SID) of 1 mini-slot or guarantee a quiet mini-slot (dead time) immediately after these
bursts and before a spreader-on interval starts.

During spreader-off bursts on S-CDMA channels when less than 128 active codes are in use, the spreader-off frame
will contain quiet mini-slots (dead time) equal to the number of inactive codes/number of codes per mini-slot, which
provides a duration equal to the number of inactive codes (128-p)*number of spreading intervals (K)*duration of a
modulation interval (T
s
).
8

6.2.12.6 Limiting the Number of Codes Assigned to a CM
In certain situations, it may be useful for a CMTS to limit the number of codes that a single CM is required to
simultaneously transmit. By doing so, the CM can divide its transmit power across a smaller number of codes than it
would otherwise, which results in a higher power per code. This can be especially useful when a population of CMs
is subject to an unusually high upstream attenuation, such that the CMs are transmitting at the maximum total
transmit power. When the value of Maximum Scheduled Codes is set less than the Number of Active Codes, the
CMTS MUST ensure that each compliant CM will not, via scheduled grants or multicast IEs with IUC=1, exceed its
assigned Maximum Scheduled Codes transmission limit in any S-CDMA frame. To accomplish this, the CMTS must
avoid scenarios that would potentially cause the CM to attempt to transmit on more codes than its Maximum
Scheduled Code limit would allow. For instance, the CMTS must manage the number of codes assigned to
contention IEs with IUC=1 in all frames. In frames where IEs with IUC=1 could not be inserted by a CMTS because
of a CM's Maximum Scheduled Codes, the CMTS MAY provide multicast IEs with IUC=2 for contention request
opportunities. CMs with Maximum Scheduled Codes enabled MUST be configurable via SNMP to control usage of

8
Revised paragraph per PHYv3.0-N-08.0701-2 on 12/23/08 by CP.
11/17/11 CableLabs
45
IEs with IUC=2 [OSSI3.0]. By default, CMs with Maximum Scheduled Codes Enabled MUST NOT use IEs with
IUC=2. Maximum Scheduled Codes MUST be equivalent to an integer number of mini-slots.
A CM MUST NOT concatenate packets beyond the size permitted by the S-CDMA Maximum Scheduled Codes, if
Maximum Scheduled Codes specified in the RNG-RSP is not 0. This is in order to reduce fragmentation overhead,
which can become significant as the number of codes reduces. A CM receiving a Maximum Scheduled Codes value
MUST be capable of fragmenting any MAC frame, including frames transmitted prior to completing the registration
process. To support 1.0 style configuration files, a CM and CMTS using the Maximum Scheduled Codes value
SHOULD support fragmentation in 1.0 mode.
If a UGS flow is requested to provide an Unsolicited Grant Size greater than the value permitted by the Maximum
Scheduled Codes value, the CMTS MUST reject the request for the UGS flow or change the CM's Maximum
Scheduled Codes value such that it would permit the UGS grants.

6.2.13 S-CDMA Framer
The S-CDMA framer maps mini-slots to spreading codes and spreading intervals by arranging them as symbols
within an S-CDMA frame. It also performs an interleaving function, to provide protection against impulse noise. The
S-CDMA framer's function of mapping mini-slots to spreading codes and spreading intervals is illustrated in Section
6.2.12, "S-CDMA Framer and Interleaver." As previously described, an S-CDMA frame is defined by the number of
spreading intervals per frame, codes per mini-slot, and number of active codes. The framer uses this information to
map the mini-slots of a transmission into frames. The framer maps complete grants so that any interleaving which is
performed is not constrained by individual mini-slot boundaries. The framer MUST align transmissions to begin and
end on mini-slot boundaries. Within a transmission, the framer numbers the symbols or bits and allocates them to
codes and spreading intervals independent of the mini-slot mapping. When using TCM encoding, the TCM encoded
symbols from the TCM encoder are split into two subsymbols consisting of the coded subsymbol which is the two
bits and the parity generated from the convolution encoder, and the uncoded subsymbol consisting of the rest of the
bits. When TCM is off, the randomizer output is treated as a continuous bit stream ignoring byte boundaries, as
specified in Section 6.2.14, "Symbol Mapping."
6.2.13.1 Subframe Definition
The S-CDMA framer performs interleaving independently of mini-slots. Interleaving is constrained by subframe
boundaries, where a subframe is a rectangular subset of an S-CDMA frame over which interleaving is performed. A
subframe is normally an integer number of Reed-Solomon code words to enhance protection from impulse noise.
Given an S-CDMA frame which is N
a
active code by K spreading intervals, a subframe is defined to be a group of R
contiguous rows, where R is an integer in the range from 1 to N
a
. A subframe is defined to exist entirely within a
single frame and does not span multiple frames. Each subframe contains R*K locations and each location holds one
symbol used for mapping and spreading. Each transmission MUST start with a new subframe. The last subframe of a
frame MUST be shortened to fit entirely within a single S-CDMA frame. The last subframe of a transmission MUST
be shortened to fit within the granted mini-slots. In both of these cases the subframe will be only R' rows instead of R
rows where R' R. Figure 618 shows a subframe consisting of R rows and K spreading intervals within an S-
CDMA frame.
46 CableLabs
11/17/11
Framelength(K)
N
u
m
b
e
r
o
f
A
c
t
i
v
e
C
o
d
e
s
(
N
a
)
R
c
o
d
e
s
SpreadingInterval
Othersubframes
Row1
Row2
Othersubframes
(K-1) 0 1

Figure 618 - Subframe structure

The parameters that define a subframe and the numbering within a subframe are codes per subframe and interleaver
step size. These two parameters are specified as part of the burst attributes and can vary between burst profiles.
These parameters determine the size of the subframe, and also how the subframe is filled with symbols. The valid
range for codes per subframe is from one to the number of active codes in use. The parameter interleaver step size is
used while putting TCM coded subsymbols and preamble symbols into the frame. Both of these types of symbols fill
in subframes first along a row, and the interleaver step size parameter indicates the spreading interval increment to
be used while filling in the symbols.
6.2.13.2 Framer Operation
The symbols entering the framer MUST be placed into the framer according to the following sets of rules. There are
two sets of rules which apply to different types of input symbols. Preamble symbols and coded TCM subsymbols
follow one set of rules, while non-TCM encoded symbols and uncoded TCM subsymbols follow the second set of
rules. The rules are specified in the following sections.
6.2.13.2.1 Rules for Preamble and Coded TCM Symbols
The CM preamble (whether TCM is on or off) and the coded TCM subsymbols MUST fill in the frame according to
the following rules:
1. The first symbol or subsymbol MUST be placed in the first spreading interval of the first row of the granted
mini-slot. In Figure 618 this would be row 1, spreading interval 0, assuming that this is the start of the first
mini-slot of the grant.
2. Subsequent symbols MUST be placed at the next available spreading interval Interleaver Step Size away from
the previous. For instance if the previous symbol was placed at spreading interval X, the next symbol is placed
at X + Interleaver Step Size.
3. If the addition of the Interleaver Step Size results in the next location being beyond the end of the frame, the
next location MUST be located modulo the frame length. For instance if J + Interleaver Step Size = K+1, then
the next location would be spreading interval 1.
11/17/11 CableLabs
47
4. If the next location is already occupied, then the spreading interval MUST be incremented by 1 until the next
unoccupied spreading interval is located. For instance if the desired location is spreading interval X and
spreading interval X is occupied, but not X+1, then X+1 would be used.
5. After filling all of the spreading intervals of a single row, the operation is repeated, starting with the next row
and step 1 above.
6. After placing all of the preamble and data symbols into the frame, the remaining symbols in the burst, as defined
by the burst boundaries of Section 6.2.5.1.1, MUST be filled with zero data symbols which will be mapped to
non-zero power.
7. Any locations that have only a TCM uncoded subsymbol MUST be filled with zero bits in the coded subsymbol
portion before mapping and spreading.
6.2.13.2.2 Rules for Uncoded Symbols and the Uncoded TCM Subsymbols
Symbols without TCM encoding and uncoded TCM subsymbols MUST fill subframes according to the following
rules.
1. The first symbol MUST be placed in the first available code of the first available spreading interval of the
subframe after the preamble has been placed into the frame. The symbols are filled from row 1 through row R
and, after filling a spreading interval, the next spreading interval is filled from row 1 through row R.
2. Uncoded symbols and the uncoded portion of TCM symbols MUST NOT be placed in the same frame location
(spreading interval, code) as a preamble symbol. For instance if there is a preamble symbol in row X, spreading
interval Y; and row (X+1), spreading interval Y is unused, the symbol should be placed into row (X+1),
spreading interval Y.
3. Subsequent symbols MUST be placed in the next available row of the first available spreading interval of the
current subframe. This causes the subframe to be filled column-wise bottom to top and then from left to right.
For instance if row 1 through row R of spreading interval X is already occupied, the next symbol would be
placed into the first available row of spreading interval X+1.
4. After completely filling a subframe, the next subframe MUST begin as specified in step 1 above.
5. The number of rows contained in the last subframe of a frame MUST be reduced to fit entirely within the frame
if there is not adequate space for a full subframe.
6. The number of rows contained in the last subframe of a grant of mini-slots MUST be reduced to fit entirely
within the granted mini-slots if there is not adequate space for a full subframe within the grant.
7. After placing all of the data symbols into the frame, the remaining symbols in the burst, as defined by the burst
boundaries of Section 6.2.5.1.1, MUST be filled with zero data symbols which will be mapped to non-zero
power.
8. Any locations that have only a TCM coded subsymbol MUST be filled with zero bits in the uncoded subsymbol
portion before mapping and spreading.
6.2.13.2.3 Subframe Example
Figure 619 below shows an example which follows the above specified rules. Each box in the figure represents a
symbol which can contain either a preamble symbol, an uncoded symbol when not using TCM, or an uncoded and
coded subsymbol when using TCM. In this example there are 9 spreading intervals in the frame, 3 rows for the
subframe, an Interleaver Step Size of 3, and the preamble is 4 symbols. Based on these parameters, the subframe
would be filled as shown. If the data is TCM encoded, the Cs would represent locations of the coded subsymbols and
the Us represent the locations of the uncoded subsymbols. If the TCM is not used, then the symbols would be placed
according to the Us only.
48 CableLabs
11/17/11
C
14
C
17
C
20
C
18
C
21
C
15
C
19
C
22
C
16
C
8
C
11
C
5
C
9
C
12
C
6
C
10
C
13
C
7
P
3
P
0
P
1
P
2
step = 3
3

r
o
w
s
frame length
U
1
U
3
U
6
U
11
U
14
U
8
U
19
U
22
U
16
U
2
U
5
U
0
U
10
U
13
U
7
U
18
U
21
U
15
U
4
U
9
U
12
U
17
U
20
C
2
C
0
C
3
C
1
C
4

Figure 619 - Symbol Numbering With and Without TCM

6.2.13.2.4 Frame Transmission
Once a frame is completed and ready for transmission, the symbols MUST be mapped and spread in spreading
interval order. This means that spreading interval 0, as shown in Figure 618, MUST be the first spreading interval
on the wire. For TCM encoded data, the coded and uncoded subsymbols from each location in the frame MUST be
combined to create complete symbols before mapping and spreading. This corresponds to creating a new symbol
where the coded portion of the symbol is C
i
and the uncoded portion is U
j
. The preamble symbols remain intact.
6.2.14 Symbol Mapping
The modulation mode is configurable via MAC messages. Differential encoded QPSK and 16-QAM are available
for TDMA channels. QPSK, 8-QAM, 16-QAM, 32-QAM, and 64-QAM are available for TDMA and S-CDMA
channels. TCM encoded QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM and 128-QAM are available for S-CDMA
channels. The symbols transmitted in each mode and the mapping of the input bits to the I and Q constellation
MUST be as defined in Table 63. In the table, x
1
represents the LSB of each of the symbol maps and x
2
, x
3
, x
4
, x
5
,
x
6
and x
7
represents the MSB for QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, and 128-QAM respectively. The
MSB MUST be the first bit in the serial data into the symbol mapper. The MSB MUST be mapped to the MSB of
the symbol map. The number of data bytes may not map into an integer number of symbols. In this case, the last
symbol MUST be padded with zero bits in the LSB locations after all data bits are processed.
All constellations are defined on a common integer grid in Figure 620. This defines each QAM symbol with 5-bit
values on each (I and Q) axis. The relative symbol amplitudes defined by the grid MUST be maintained across all
constellations. Different constellations may be used, for example, in different burst profiles, in preamble and data
symbols within the same burst, and in modulating different spreading codes within a frame.
In Figure 620, E
av
denotes the average constellation energy for equally likely symbols. For each constellation the
integer values of E
av
and differences in dB compared to 64-QAM, G
const
, are given. The QPSK0 constellation is
employed for low-power preamble and all QPSK data symbols. Use of QPSK1 is restricted to high-power preamble
symbols.
11/17/11 CableLabs
49
Table 63 - I/Q Mapping
QAM Mode Input bit Definitions
QPSK x
2
x
1
8-QAM x
3
x
2
x
1
16-QAM x
4
x
3
x
2
x
1
32-QAM x
5
x
4
x
3
x
2
x
1
64-QAM x
6
x
5
x
4
x
3
x
2
x
1
128-QAM x
7
x
6
x
5
x
4
x
3
x
2
x
1

The upstream symbol constellations MUST be as shown in Figure 620.
The upstream QPSK Gray-coded and differential symbol mapping MUST be as shown in Figure 621.
The upstream 8-QAM symbol mapping MUST be as shown in Figure 622.
The upstream 16-QAM Gray-coded symbol mapping MUST be as shown in Figure 623.
The upstream 16-QAM differential symbol mapping MUST be as shown in Figure 623.
The upstream 32-QAM symbol mapping MUST be as shown in Figure 624.
The upstream 64-QAM Gray-coded symbol mapping MUST be as shown in Figure 625.
The TCM symbol mapping used for S-CDMA is shown in Figure 626 through Figure 628.
The upstream QPSK TCM symbol mapping MUST be as shown in Figure 626.
The upstream 8-QAM TCM symbol mapping MUST be as shown in Figure 626.
If differential quadrant encoding is enabled, then the currently-transmitted symbol quadrant is derived from the
previously transmitted symbol quadrant and the current input bits via Table 64. If differential quadrant encoding is
enabled, the upstream PMD sublayer MUST apply these differential encoding rules to all transmitted symbols
(including those that carry preamble bits). Differential quadrant encoding is only available for QPSK and 16-QAM
on TDMA channels. In Table 64, I(1)Q(1) refers to x
2
x
1
and x
4
x
3
from Table 63 for QPSK and 16-QAM cases
respectively.
50 CableLabs
11/17/11
Table 64 - Definition of Differential Quadrant Coding
Current Input
Bits I(1) Q(1)
Quadrant
Phase Change
MSBs of Previously
Transmitted Symbol
MSBs for Currently
Transmitted Symbol
00 0 11 11
00 0 01 01
00 0 00 00
00 0 10 10
01 90 11 01
01 90 01 00
01 90 00 10
01 90 10 11
11 180 11 00
11 180 01 10
11 180 00 11
11 180 10 01
10 270 11 10
10 270 01 11
10 270 00 01
10 270 10 00

11/17/11 CableLabs
51
QPSK1: E
av
= 288 (G
const
=+2.34 dB)
-12 0 +12
QPSK0: E
av
= 128 (G
const
=-1.18 dB rel to 64 QAM)
+15
+12
+8

0

-8
-12
-15
-15 -8 +8 +15
Imaginary part
(quadrature)
Real part
(in-phase)
16 QAM-SQ: E
av
= 160 (G
const
=-0.21 dB)
+12

+4

-4

-12
8 QAM-DS: E
av
= 160 (G
const
=-0.21dB)
+12

+4

-4

-12
-12 -4 +4 +12
-12 -4 +4 +12
64 QAM-SQ: E
av
= 168 (G
const
= 0 dB)
+14

+10

+6

+2
-2

-6

-10

-14
32 QAM-DS: E
av
= 168 (G
const
=0 dB)
+14

+10

+6

+2
-2

-6

-10

-14
-14 -10 -6 -2 +2 +6 +10 +14
-14 -10 -6 -2 +2 +6 +10 +14
128 QAM-DS: E
av
= 170 (G
const
= +0.05 dB)
+15
+13
+11
+9
+7
+5
+3
+1
-1
-3
-5
-7
-9
-11
-13
-15
-15-13-11 -9 -7 -5 -3 -1 +1 +3 +5 +7+9+11+13+15
QPSK1
QPSK0

Figure 620 - Symbol Constellations

52 CableLabs
11/17/11
Q
I
Label bits x
2
x
1
(I, Q)
01
00 10
11

Figure 621 - QPSK Gray and Differential Symbol Mapping

Q
I
Label bits x
3
x
2
x
1
110
000
001
111
101
011
010
100
Gray-code
violation
2
2
2

Figure 622 - 8-QAM Symbol Mapping

11/17/11 CableLabs
53
Q
I
Label bits x
4
x
3
x
2
x
1
1111
0100 1100 1110
0111 0101
0110
1101
0000 0010 1010 1000
0011 0001 1001 1011
(a) Gray-Coded
Mapping (I, Q, I, Q)
Q
1111
0100 1100 1110
0111 0110
0101
1101
0000 0010 1001 1000
0011 0001 1010 1011
(b) Mapping for Differential
Encoding


Q
I
Label bits x
5
x
4
x
3
x
2
x
1
2
2
2
2
2
2
10111
11001 01011
11000 01111 01001 00110
00111 01000 00010 01101
00101 10000
00011
10101 00000
10100
00001
01100
10011
11100 10001
00100
10010 11010
01010
11111
10110
11101
11110
11011
01110
Gray-code violation
2

9

9
Revised graphic per PHYv3.0-N-08.0701-2 on 12/23/08 by CP.
54 CableLabs
11/17/11
Q
I
Label bits x
6
x
5
x
4
x
3
x
2
x
1
(I, Q, I, Q, I, Q)
011100 011110 010110 010100
011101 011111 010111 010101
011001 011011 010011 010001
011000 011010 010010 010000
001000 001010 000010 000000
001001 001011 000011 000001
001101 001111 000111 000101
001100 001110 000110 000100
110100 110110 111110 111100
110101 110111 111111 111101
110001 110011 111011 111001
110000 110010 111010 111000
100000 100010 101010 101000
100001 100011 101011 101001
100101 100111 101111 101101
100100 100110 101110 101100


Binarylabelsx
2
x
1
01
11 10
00
Binarylabelsx
3
x
2
x
1
101
011
001
110
010
000
100
111

Figure 626 - QPSK and 8-QAM TCM Symbol Mapping

11/17/11 CableLabs
55
Octal labels o
2
o
1
(= bin. labels x
4
x
3
x
2
x
1
)
13
11 10
12 07 16
04 01 00 15
17 02 03 06
05 14
Subset B0
21
26
34 16 24
23 31 13
Subset B1
15 07 05 37
10 02 00 32
33 01 03
11
36 04
06 14
25 17 35 27
20 12
30 22
Octal labels o
2
o
1
(= bin. labels x
5
x
4
x
3
x
2
x
1
)

Figure 627 - 16-QAM and 32-QAM TCM Symbol Mapping

64 QAM: octal labels o
2
o
1

(= bin. labels x
4
x
3
x
2
x
1
)
Subset B0
62 43 46 77 23 57 26 42
16 27 32 13 07 73 12 76
02 63 66 17 03 37 06 22
36 47 52 33 67 53 72 56
54 51 70 65 34 31 50 45
20 35 04 01 00 15 64 61
74 71 10 05 14 11 30 25
40 55 24 21 60 75 44 41
Subset B1
104
103 111 173 141 043 051 133 101
106 114 176 144 046 054 136
112
135 127 165 157 075 067 125 117
130 122 160 152 070 062 120
174
053 061 023 031 013 021 163 171
056 064 026 034 016 024 166
142
045 077 015 007 005 037 155 147
040 072 010 002 000 032 150
102
105 137 055 047 145 177 115 107
100 132 050 042 140 172 110
044
143 151 033 001 003 011 073 041
146 154 036 004 006 014 076
052
175 167 025 017 035 027 065 057
170 162 020 012 030 022 060
134
113 121 063 071 153 161 123 131
116 124 066 074 156 164 126
128 QAM: octal labels o
3
o
2
o
1

(= bin. labels x
7
x
6
x
5
x
4
x
3
x
2
x
1
)

Figure 628 - 64-QAM and 128-QAM TCM Symbol Mapping
10

6.2.15 S-CDMA Spreader
The basis of signal transmission with S-CDMA is direct-sequence spread-spectrum modulation. S-CDMA employs a
family of orthogonal digital code words, called spreading codes, to simultaneously transmit up to 128 modulation
symbols. In each spreading interval, a vector Pk is transmitted such that:
P
k
= S
k
*C,

10
Revised graphic per PHYv3.0-N-08.0701-2 on 12/23/08 by CP.
56 CableLabs
11/17/11
where S
k
is a vector, [s
k,127
, s
k,126
,..., s
k,0
], of modulation symbols on the integer grid of Section 6.2.14 to be
transmitted in spreading interval k, and C is a matrix:
(
(
(
(
(
(
=
(
(
(
(
(
=
1 1 1 1 1
1 ...
1 ... ... ... ...
1 ...
1 ...
...
... ... ... ...
...
...
1 126 127
3 1 2
2 127 1
0 0, 126 0, 127 0,
0 126, 126 126, 127 126,
0 127, 126 127, 127 127,
x x x
x x x
x x x
C C C
C C C
C C C
C
where the rows of C are the 128 spreading codes such that code i = [c
i,127
, c
i,126
,..., c
i,0
]. The matrix C may be written
in shorthand form in terms of its rows as follows:
code 127
code 126
...
code 2
code 1
code 0
C
(
(
(
(
=
(
(
(
(
(

The result of the spreading operation is the transmission vector P
k
which has 128 elements, (p
k,127
, p
k,126
,..., p
k,0
),
where each element is transmitted at the signaling rate, with element p
k,0
transmitted first in time. The first element S
0

into the spreader is defined as follows. As a point of reference, for 128 allocated codes, and considering the first
column of the framer (k = 0), S
0
is the first symbol in time to enter the framer, occupies the lower left element of the
framer, and is the first element into the spreader.
The set of orthogonal codes used for the spreading operation is quasi-cyclic and consists of values which are either
+1 or -1. Code 0 consists of 128 elements, each of which has value +1. For each of the other spreading codes, code i,
the element c
i,0
is -1 and the remaining elements are obtained by a cyclic shift of a sequence x as is shown in the
above matrix in this section.
The sequence x
i
is defined such that the elements corresponding to the following set of indices are equal
to -1:
{2 3 4 5 6 7 9 10 11 13 16 17 18 19 20 21 25 26 28 30 31 33 34 35 37 39 40 41 49 51 52 55 56 59 60 61 65 66 67
69 72 73 74 77 78 79 81 84 90 92 94 97 100 101 103 106 109 110 111 114 117 119 121};
The remaining elements of code 1 have a value of +1.
Each code i is obtained by cyclic shift to the left (in the direction of increasing indices) of code (i-1), where the
element c
i,0
has a value of -1 and does not take part in the cyclic shift.
Although each code is defined to have equal power, the spread symbols may have slightly unequal power since the
symbols at the input to the spreader have varying values of E
av
according to the integer symbol grid of Section
6.2.14.
If a CM has not been assigned to use a particular code i at a spreading time interval k, then in its computation of its
transmission vector P
k
, it will set s
k,i
to numerical zero. The assignment of codes to the CM is performed by the
framer as it assigns a burst of symbols a particular order in the two-dimensional space of codes and time. This
symbol sequencing is described in detail in Section 6.2.13.
11/17/11 CableLabs
57
The I and Q components of the symbols are spread using the same spreading code.
In the matrix multiple of the equation above, and subsequent CM processing prior to the D/A, there is an essential
clipping operation wherein, as an example, filtered (pulse-shaped) elements of P
k
in excess of some vendor-specific
absolute value are clipped (retaining complex angle) to this absolute value. This non-linear operation, deviating from
the equation above and the subsequent linear processing prior to the D/A, is essential for meeting spurious emission
and MER requirements safely and efficiently, while operating at the highest CM average transmit power levels (see
Table 612, "User Unique Burst Parameters").
6.2.15.1 Code Hopping
Code hopping refers to a systematic re-ordering of the rows of the original spreading matrix C, so that at each
spreading interval k, a new code matrix C
k
is produced. A pseudo-random number generator determines a cyclic shift
of a subset of the rows of the original matrix C. Two code hopping modes are defined: Code Hopping Mode 1 is
used exclusively with Selectable Active Codes Mode 1; Code Hopping Mode 2 is used exclusively with Selectable
Active Codes Mode 2.
In S-CDMA mode, the CM MUST support Code Hopping Mode 1. In S-CDMA mode, the CM SHOULD support
Code Hopping Mode 2. The CMTS MAY support code hopping.
11

6.2.15.1.1 Code Hopping Mode 1
In Code Hopping Mode 1, when the number of active codes N
a
< 128, the cyclic codes are hopped (cyclically
shifted), while code 0, the all 1s code, remains fixed at the bottom of the matrix. The unused codes are located in the
(128 N
a
) bottom rows of the hopped matrix C
k
. Thus, for N
a
< 127, the set of unused codes changes at each
spreading interval. When N
a
= 128, all codes are hopped, including code 0. The hopped spreading matrix is defined
by:
( ) ( ) ( )
( ) ( ) ( )
( ) ( ) ( ) (
(
(
(
(
=
0 , 0 k, f 126 , 0 k, f 127 , 0 k, f
0 , 126 k, f 126 , 126 k, f 127 , 126 k, f
0 , 127 k, f 126 , 127 k, f 127 , 127 k, f
k
...
... ... ... ...
...
...
C C C
C C C
C C C
C
where,
( )
( )
( )
modulo 128 _ ( ) ,128), 128, 0 127
,
modulo 126 _ ( ) ,127) 1, 128, 1 127
a
a
lfsr out k i N i
f k i
lfsr out k i N i
+ =
=

+ + <

The matrix elements c
i,j
are the elements of the original code matrix C
k indicates the spreading interval;
i indicates the row in the matrix;
lfsr_out is a pseudo-random number described in Section 6.2.15.1.3.

11
58 CableLabs
11/17/11
6.2.15.1.2 Code Hopping Mode 2
In Code Hopping Mode 2, the active codes are hopped, while the unused codes remain fixed at the bottom of the
matrix. A horizontal line in the following equation shows the division between active and unused codes. The hopped
spreading matrix is defined by:
( ) ( ) ( )
( ) ( ) ( )
( ) ( ) ( )
(
(
(
(
(
(
(
(
(
(
(
(
(
(
=
0 u(0), 126 u(0), 127 u(0),
0 1), - Na - u(128 126 1), - Na - u(128 127 1), - Na - u(128
0 , Na - 128 k, f 126 , Na - 128 k, f 127 , Na - 128 k, f
0 , 126 k, f 126 , 126 k, f 127 , 126 k, f
0 , 127 k, f 126 , 127 k, f 127 , 127 k, f
k
...
... ... ... ...
...
...
... ... ... ...
...
...
C C C
C C C
C C C
C C C
C C C
C
where,
f(k,i)= active_code_list [ modulo(2*N
a
-128-hop_number(k)+i, N
a
) ]
128-N
a
i 127
The matrix elements c
i,j
are the elements of the original code matrix C:
k indicates the spreading interval;
i indicates the row in the matrix;
N
a
is the number of active codes;
active_code_list is the list of active codes, in increasing order, with each element in the range 0 to 127;
u is the list of unused codes, in increasing order, with each element in the range 0 to 127;
hop_number is a pseudo-random number described in Section 6.2.15.1.3.
The following is a numerical example of Selectable Active Codes Mode 2 and Code Hopping Mode 2. Figure 616
shows N
a
= 124 active codes with unused code list u = {0, 1, 5, 125}. Thus, active_code_list = {2, 3, 4, 6, 7, 8, 9, 10,
..., 123, 124, 126, 127}, as annotated at the left of the figure; active_code_list(0) = 2, active_code_list(123) = 127,
u(0) = 0, and u(3) = 125. Assume that at a given spreading interval k, the pseudo-random hop generator returns a
value of hop_number = 0. (This results in the same matrix that would occur with code hopping turned off). The
spreading matrix C
k
consists of the unused codes at the bottom and active codes in increasing order on top:
11/17/11 CableLabs
59
code 127
code 126
code 124
code 123
...
code 8
code 7
code 6
code 4
code 3
code 2
code 125
code 5
code 1
code 0
k
C
(
(
(
(
(
(
(
(
(
(
(
(
(
=
(
(
(
(
(
(
(
(
(
(
(
(
(

At another spreading interval, if hop_number = 3, the active-code rows are cyclically shifted vertically by 3 rows,
while the unused codes remain fixed, and the spreading matrix C
k
becomes:
code 123
code 122
code 121
code 120
...
code 9
code 8
code 7
code 6
code 4
code 3
code 2
code 127
code 126
code 124
code 125
code 5
code 1
code 0
k
C
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
=
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(

60 CableLabs
11/17/11
6.2.15.1.3 Code hopping generator
The pseudo-random number generator, which determines the spreading matrix reordering during code hopping,
utilizes the linear-feedback shift register (LFSR), shown in Figure 629. In order to align the CM's code-hopping
pseudo-random sequence with that of the CMTS, the pseudo-random generator must output the following value at
the first spreading interval of each frame:
lfsr_out(frame_number * spreading_interval_per_frame)
where lfsr_out(k) is the value of lfsr_out, after k shift register updates, following the code hopping seed load into the
LFSR. The bits are defined as lfsr_out 7:1 = s7:1, where s is the shift register contents, and lfsr_out bit 7 is
considered the MSB. In Code Hopping Mode 1, a shift register update equates to one shift of the LFSR. In Code
Hopping Mode 2, a shift register update equates to 15 shifts of the LFSR.
The description of the frame counter and the procedures for its synchronization are contained in Section 6.2.12.2,
"Mini-slot Numbering." At this reset, a 15-bit initialization value (seed) is loaded into the shift register and is used at
the first spreading interval. The 15-bit seed value is configured in response to the Upstream Channel Descriptor
message from the CMTS.
At each subsequent spreading interval k, the LFSR is updated. The code hopping mechanism (LFSR and spreading
interval index k) is advanced, every spreading interval (128 modulation intervals), in both spreader-on and spreader-
off frames.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
Initializationvalue(seed)
NumberofactivecodesNa
Divide
by
2^15
floor
Codehoppingmode1:
clockoncebeforeeach
use
Codehoppingmode2:
clock15timesbefore
eachuse
wordassemble
y15:1=[s7:1,s15:8]
s1 s15
lfsr_out=s7:1 s15:8
y
y15=MSB
y1=LSB
lfsr_out
(outputforcodehopping
mode1)
hop_number
(outputforcodehopping
mode2)

Figure 629 - Code Hopping Random Number Generator

6.2.15.1.3.1 Code Hopping Mode 1 pseudo-random number generation
In Code Hopping Mode 1, the LFSR is updated by clocking it once before each use on each spreading interval. The
output lfsr_out is used to compute the spreading matrix indices, as given by the equation for Code Hopping Mode 1
above.
11/17/11 CableLabs
61
6.2.15.1.3.2 Code Hopping Mode 2 pseudo-random number generation
In Code Hopping Mode 2, the LFSR is updated by clocking it 15 times before use on each spreading interval,
thereby flushing it to reduce the correlation between hops. The LFSR contents are then scaled to produce an
approximately uniform distributed pseudo-random hop number in the range 0 to N
a
- 1. The scaling consists of the
following operation:
15
_
2
a
N y
hop number floor
| |
=
|
\ .

In this equation, N
a
is the number of active codes, and y is the result of assembling the output of the LFSR into a 15-
bit word as follows: Bits 7:1 of the LFSR are mapped to the most significant 7 bits of y: The remaining bits 15:8 of
the LFSR are used as the least significant 8 bits of y. That is, y15:1 = [s7:1, s15:8], where y15 is considered the
MSB. The output hop_number is used to compute the spreading matrix indices, as given by the equation, for Code
Hopping Mode 2 above.
Full precision MUST be maintained in the multiplication of N
a
*y. The product MUST then be truncated, discarding
the 15 LSBs, as shown in the figure. Or, equivalent processing may be used that produces the same hop_number
output under all input conditions. The use of the full 15 bits of the LFSR provides a dithering effect which prevents
"clumping" in the hopping distribution.
6.2.16 Transmit Pre-Equalizer
A transmit pre-equalizer of a linear equalizer structure, as shown in Figure 630, MUST be configured for each
upstream channel used by the CM in response to the Ranging Response (RNG-RSP) message transmitted by the
CMTS.
There are two modes of operation for the pre-equalizer of a CM: DOCSIS 1.1 mode, and DOCSIS 2.0 pre-
equalization mode: In DOCSIS 1.1 mode, the CM MUST support a (T)-spaced equalizer structure, with 8 taps. The
pre-equalizer MAY have 1, 2, or 4 samples per symbol, with a tap length longer than 8 symbols. In DOCSIS 1.1 pre-
equalization mode, for backwards compatibility, the CMTS MAY support fractionally spaced equalizer format (T/2
and T/4). In DOCSIS 2.0 pre-equalization mode, the pre-equalizer MUST support a symbol (T)-spaced equalizer
structure with 24 taps.
In DOCSIS 1.x-only logical channels (type 1 logical channels [MULPI]), the CM and the CMTS MUST use
DOCSIS 1.1 pre-equalization mode.
In DOCSIS 2.0-only or DOCSIS 3.0-only logical channels (type 3 or type 4 logical channels [MULPI]), the CM and
the CMTS MUST use DOCSIS 2.0 pre-equalization mode.
In DOCSIS 1.x/2.0 mixed logical channels (type 2 logical channels [MULPI]), the CM and the CMTS MUST use
DOCSIS 1.1 pre-equalization mode from initial ranging until DOCSIS 2.0 mode is enabled or until Multiple
Transmit Channel mode is enabled in the registration process (if either is enabled). The CM and CMTS MUST use
DOCSIS 2.0 pre-equalization mode after DOCSIS 2.0 is enabled for the CM. In the case that a CM is placed into
Multiple Transmit Channel mode, the CM and CMTS use DOCSIS 2.0 pre-equalization for all bursts on type 2
logical channels whether the burst descriptor in the UCD for the transmitted burst is a Type 4 or Type 5 burst
descriptor.
62 CableLabs
11/17/11
+
2 3
I-QInput
z
-1
z
-1
z
-1
z
-1
z
-1
z
-1
+ + + + +
... ...
... ...
F
2
F
1
F
4
F
3
F
17
F
18
F
24
Equalizer
Output

Figure 630 - Transmit Pre-Equalizer Structure

The RNG-RSP MAC message carries the CM transmit equalization information, and may instruct the CM to either
convolve the equalizer coefficients, or (in DOCSIS 2.0 pre-equalization mode only) load them directly. When the
CM is instructed to convolve the transmit equalizer coefficients, it MUST convolve the coefficients sent by the
CMTS in the RNG-RSP with the existing coefficients, to get the new coefficients. After convolving, the CM MUST
truncate the convolution result, such that 24 taps (8 taps in DOCSIS 1.1 pre-equalization mode) remain after the
truncation, with the main tap located at the tap designated by the last RNG-RSP received by the CM. The operation
of the convolution is formulized by the following equation:
( )
( )
24 ... 1 ,
1
1 1
1 1
1
^
1 , 24 min
24 . 1 max
1
= =
+
+ +
+ +
+ +
+
+ =
+
+
n F F F
m
m m m
m m m
m m l k
L L n L
L L n L k
m
L L k n
m
n
where:
m
n
F are the coefficients prior to the convolution;
1 m
n
F
+
are the coefficients after the convolution;
n
F
^
are the coefficients sent from the CMTS;
m
L is the main tap location prior to the convolution;
1 + m
L is the main tap location after the convolution as dictated by the CMTS.
In DOCSIS 2.0 pre-equalization mode, the CMTS MAY instruct the CM to load the transmit pre-equalizer
coefficients. When the CM is instructed to load the transmit equalizer coefficients, it MUST load the coefficients
sent by the CMTS into the pre-equalizer coefficients after proper normalization, if necessary.
In DOCSIS 1.x-only logical channels, in response to an initial ranging request and periodic ranging requests prior to
CM registration, when the CMTS sends the pre-equalizer coefficients, the CMTS MUST compute and send them
with an equalizer length of 8 and in T-spaced format, where T is the modulation interval. After registration, the
CMTS MAY use a fractionally spaced equalizer format (T/2- or T/4-spaced), with a longer tap length, to match the
CM pre-equalizer capabilities that the CMTS learned from the REG-REQ message modem capabilities field.
11/17/11 CableLabs
63
In DOCSIS 2.0-only or DOCSIS 3.0-only logical channels, the CMTS MUST compute and send the pre-equalizer
coefficients with an equalizer length of 24, and in T-spaced format, at all times.
In DOCSIS 1.x/2.0 mixed logical channels, in response to an initial ranging request and periodic ranging requests
prior to CM registration, when the CMTS sends the pre-equalizer coefficients, the CMTS MUST compute and send
them with an equalizer length of 8 and in T-spaced format. After registration, if the DOCSIS 1.1 pre-equalization
mode is enabled, the CMTS MAY use a fractionally spaced equalizer format (T/2- or T/4-spaced), with a longer tap
length, to match the CM pre-equalizer capabilities that the CMTS learned from the REG-REQ message modem
capabilities field. If DOCSIS 2.0 pre-equalization mode or Multiple Transmit Channel mode is enabled for the CM,
the CMTS MUST use a T-spaced equalizer structure with 24 taps. If the first update of the pre-equalizer after the
activation of DOCSIS 2.0 pre-equalization mode uses "convolve" mode, the CM MUST zero-pad the existing 8-tap
filter to a 24-tap filter, and convolve, according to the rules above.
Prior to making an initial ranging request, and whenever the upstream channel frequency or upstream channel
modulation rate changes, the CM MUST initialize the coefficients of the pre-equalizer to a default setting in which
all coefficients are zero, except the real coefficient of the first tap (i.e., F
1
). Whenever the main location is changed,
the CM, not the CMTS, MUST compensate for the delay (ranging offset), due to a shift from the previous main tap
location, to a new main tap location of the equalizer coefficients, sent by the CMTS (in both "convolve" and "load"
operations). The pre-equalizer coefficients are then updated through the subsequent ranging process (unicast initial
ranging and periodic ranging).
In DOCSIS 1.1 pre-equalization mode, the CMTS MUST NOT move the main tap location during periodic ranging.
In DOCSIS 1.1 pre-equalization mode, the CMTS MUST NOT instruct the CM to load the transmit equalizer
coefficients.
In DOCSIS 2.0 pre-equalization mode, the CMTS MAY move the main tap location during unicast initial ranging or
periodic ranging.
Equalizer coefficients may be included in every RNG-RSP message, but typically they only occur when the CMTS
determines that the channel response has significantly changed. The frequency of equalizer coefficient updates in the
RNG-RSP message is determined by the CMTS.
The CM MUST normalize the transmit equalizer coefficients in order to guarantee proper operation (such as not to
overflow or clip). The CM MUST NOT change its target transmit power due to gain or loss of the new coefficients
in both "convolve" and "load" operations. The target power is defined in Section 6.2.19, "Transmit Power
Requirements".
In DOCSIS 1.1 mode, if the CM equalizer structure implements the same number of coefficients as assigned in the
RNG-RSP message, then the CM MUST NOT change the location of the main tap in the RNG-RSP message. If the
CM equalizer structure implements a different number of coefficients than defined in the RNG-RSP message, the
CM MAY shift the location of the main tap value. The CM MUST adjust its ranging offset, in addition to any
adjustment in the RNG-RSP message, by an amount that compensates for the movement of the main tap location.
6.2.17 Spectral Shaping
The CM upstream transmitter MUST approximate a Nyquist square-root raised-cosine pulse-shaping filter with roll-
off factor alpha =0.25. For operation with a DOCSIS 3.0 CMTS (unless the CMTS is operating as a pre-3.0 DOCSIS
CMTS), the CM -30 dB transmitted bandwidth MUST NOT exceed the Channel Width values in Table 65. For
backward compatibility with a CMTS operating as a pre-3.0-DOCSIS CMTS, the CM -30 dB transmitted bandwidth
MUST NOT exceed the Channel Width values in Table 66. The Channel Width values are given analytically by:
ChannelWidth = ModulationRate*(1+alpha).
64 CableLabs
11/17/11

Table 65 - Maximum Channel Width for Operation with a DOCSIS 3.0 CMTS
Modulation Rate (kHz) Channel Width (kHz)
1280 1600
2560 3200
5120 6400

Table 66 - Maximum Channel Width for Operation with a pre-3.0-DOCSIS CMTS
Modulation Rate
(kHz)
Channel Width (kHz) Applicability
160 200 CM MAY support
1280 1600 CM MUST support

6.2.17.1 Upstream Frequency Agility and Range
The CM MUST support operation over the Standard Upstream Frequency Range of 5-42 MHz.
Additionally, the CM MAY support selectable operation over a) the Standard Upstream Frequency Range of 5-42
MHz, or b) the Extended Upstream Frequency Range of 5-85 MHz, edge to edge.
Offset frequency commands MUST be supported per Table 612.
6.2.17.2 Spectrum Format
The upstream modulator MUST provide operation with the format s(t) = I(t)*cos(t) - Q(t)*sin(t), where t denotes
time and denotes angular frequency.
11/17/11 CableLabs
65
6.2.18 Relative Processing Delays
The CM MAP processing delay is the time provided between arrival of the last bit of a MAP message at a CM and
the effectiveness of this MAP. During this time, the CM should process the MAP message and fill its interleavers (or
its framer, in S-CDMA mode) with encoded data. The CMTS MUST transmit the MAP message early enough to
allow the CM MAP processing delay, as specified below.
For a CM in MTC mode, the CM MAP processing delay, D
p
, is given by the equations:
12

=
+ =
0 ,
0 ,
12 . 5
600
r r
r r r
p
I B
I N I
M
s
M
D

where M is the number of elements in the CM interleavers (in the case of TDMA), or framer (in the case of S-
CDMA). If there is no upstream interleaving or S-CDMA framing, M = 0. Note that in the above equations, the
values for B
r
and I
r
*N
r
are taken to be the maximum from all of the specified burst types in a particular UCD.
In S-CDMA mode, M = 128(K+1), where K is the number of spreading intervals per frame. This is the time required
for processing an S-CDMA frame, plus an extra spreading interval. For example, in the case of K = 32, which
corresponds to the maximum framer size, the CM MAP processing time is 1425 s.
For a CM not operating in MTC mode, the CM MAP processing delay, Dp, is given by the equation with M being
equivalent to what is specified above:
sec
12 . 5
200
M
D
p
+ =
Note: The CM MAP processing delay does not include downstream FEC de-interleaving delay.
Note: The "effectiveness of the MAP" relates to the beginning of the burst frame at the RF output of the
CM. In the S-CDMA mode, "effectiveness of the MAP" relates to the beginning (at the RF output
of the CM) of the first spreading interval of the S-CDMA frame which contains the burst.
6.2.19 Transmit Power Requirements
13

The following requirements apply with Multiple Transmit Channel mode enabled. Requirements with Multiple
Transmit Channel mode disabled are addressed in Section 6.2.19.4 below.
The CM MUST support varying the amount of transmit power. Requirements are presented for 1) range of reported
transmit power per channel; 2) step size of power commands; 3) step size accuracy (actual change in output power
per channel compared to commanded change); and 4) absolute accuracy of CM output power per channel. The
protocol by which power adjustments are performed is defined in [MULPI]. Such adjustments by the CM MUST be
within the ranges of tolerances described below. A CM MUST confirm that the transmit power per channel limits are
met after a RNG-RSP is received for each of the CM's active channels that is referenced. An active channel for a CM

12
Revised per PHYv3.0-N-06.0288-1 by GO on 10/19/06; revised equation per PHYv3.0-N-06.0340-1 by GO on 1/12/07.
13
Section modified per PHYv3.0-N-07.0570-2 on 1/25/08 and per PHYv3.0-N-08.0611-3 on 4/10/08 by KN.
66 CableLabs
11/17/11
is defined as any channel for which the CM has been given a grant, which it will use for channel initialization, or
ranging, or any channel for which the CM is "ranged." The set of "active channels" is also called the Transmit
Channel Set. (There will be some mechanism by which a CMTS can command a CM to deactivate an active channel,
for the purpose of either reducing the number of active channels on that CM or for the purpose of adding a different
active channel as a replacement upstream channel for that CM.) Note that the set of channels actually bursting
upstream from a CM is a subset of the active channels on that CM; often one or all active channels on a CM will not
be bursting, but such quiet channels are still "active channels" for that CM.
Transmit power per channel is defined as the average RF power in the occupied bandwidth (channel width)
transmitted in the data symbols of a burst, assuming equal likely QAM symbols, measured at the F-connector of the
CM. Total transmit power is defined as the sum of the transmit power per channel of each channel transmitting a
burst at a given time. Maximum and minimum transmit power per channel level requirements refer to the CM's target
transmit power per channel levels, defined as the CM's estimate of its actual transmit power per channel. The actual
transmitted power per channel MUST be within 2 dB of the target power. The target transmit power per channel
MUST be variable over the range specified in Table 612. The target transmit power per channel MAY be variable
over a range that extends above the maximum levels specified in Table 67, Table 68, and Table 69. The CM
MUST NOT exceed a target transmit power level per channel of 61 dBmV. Note that all fidelity requirements
specified in Section 6.2.22 still apply when the CM is operating over an extended transmit power range.
The CM communicates its ability to transmit above the maximum levels defined in Table 67, Table 68, and Table
69 to the CMTS via a modem capability encoding as defined in [MULPI]. When the CM indicates that it supports
the extended range and the CMTS confirms this capability, for each combination of the number of channels in the
Transmit Channel Set and constellation the CM and CMTS MUST use the greater of the value from the CMs
advertised capability encoding and the default value of P
max
specified in Table 67, Table 68, and Table 69 as P
max

for that particular combination. When the CM indicates that it supports the extended range and the CMTS disables
this capability, the CM and CMTS MUST use the default values of P
max
specified in Table 67, Table 68, and
Table 69.
14

With Multiple Transmit Channel mode enabled, let P
load
= P
hi
_ P
r
, for each channel, using the definitions for P
hi
and
P
r
in the following subsections of 6.2.19. The channel corresponding to the minimum value of P
load
is called the
highest loaded channel, and its value is denoted as P
load_1
, in this specification even if there is only one channel in the
Transmit Channel Set. A channel with high loading has a low P
load
_
n
value (but not less than 0); the value of P
load_n
is
analogous to an amount of back-off for an amplifier from its max power output. A channel has lower power output
when that channel has a lower loading (more back-off) and thus a higher value of P
load_n
. Note that the highest loaded
channel is not necessarily the channel with the highest transmit power, since a channels max power depends on the
modulations it is supporting in its burst profiles. The channel with the second lowest value of P
load
is denoted as the
second highest loaded channel, and its loading value is denoted as P
load_2
; the channel with the third lowest value of
P
load
is the third highest loaded channel, and its loading value is denoted as P
load_3
; and the channel with the fourth
lowest loading value is denoted as the fourth highest loaded channel, and its loading value is P
load_4
(in the case of
fewer than 4 channels in the Transmit Channel Set, P
load_n
will only be valid for the n active upstream channels).
P
load_min_set
defines the upper end of the Dynamic Range Window for the CM with respect to P
hi
for each channel.
P
load_min_set
will limit the maximum power possible for each active channel to a value less than P
hi
when P
load_min_set
is
greater than zero. P
load_min_set
is a value commanded to the CM from the CMTS when the CM is given a TCC in
registration and RNG-RSP messages [MULPI]. P
load_min_set
, P
load_n
, P
hi_n
, P
r_n,
etc., are defined only when Multiple
Transmit Channel mode is enabled.
15

The CMTS MUST provide P
load_min_set
such that P
hi_n
P
load_min_set
P
low_n
for each active channel, or equivalently,
0 P
load_min_set
P
hi_n
P
low_n.

14
Revised per PHYv3.0-N-10.0942-5 on 9/21/10 by JB.
15
Revised this paragraph per PHYv3.0-N-06.0340-1 by GO on 1/11/07.
11/17/11 CableLabs
67
CMTS commands adjusting P
r_n
MUST be consistent with the P
load_min_set
previously assigned to the CM and with the
following limits:
P
load_min_set
P
hi_n
P
r_n
P
load_min_set
+ 12 dB
and the equivalent:
P
hi_n
(P
load_min_set
+ 12 dB) P
r_n
P
hi_n
P
load_min_set
16

A value is computed, P
low_multi_n
, for each channel in the Transmit Channel Set, which sets the lower end of the
transmit power Dynamic Range Window for that channel, given the upper end of the range (which is determined by
P
load_min_set
).
17

P
low_multi_n
= P
hi_n
P
load_min_set
12 dB.
The effect of P
low_multi_n
is to restrict the dynamic range required (or even allowed) by a CM across its multiple
channels, when operating with multiple active channels.
18

When the CMTS sends a new value of P
load_min_set
to the CM, there is a possibility that the CM will not be able to
implement the change to the new value immediately, because the CM may be in the middle of bursting on one or
more of its upstream channels at the instant the command to change P
load_min_set
is received at the CM. Some amount
of time may elapse before the CMTS grants global reconfiguration time to the CM. Similarly, commanded changes
to P
r_n
may not be implemented immediately upon reception at the CM if the n
th
channel is bursting. Commanded
changes to P
r_n
may occur simultaneously with the command to change P
load_min_set
. The CMTS SHOULD NOT issue
a change in P
load_min_set
after commanding a change in P
r_n
until after also providing a sufficient reconfiguration time
on the n
th
channel. The CMTS SHOULD NOT issue a change in P
load_min_set
after commanding a prior change in
P
load_min_set
until after also providing a global reconfiguration time for the first command. Also, the CMTS SHOULD
NOT issue a change in P
r_n
a) until after providing a global reconfiguration time following a command for a new
value of P
load_min_set
and b) until after providing a sufficient reconfiguration time on the n
th
channel after issuing a
previous change in P
r_n
. In other words, the CMTS is to avoid sending consecutive changes in P
r_n
and/or P
load_min_set

to the CM without a sufficient reconfiguration time for instituting the first command. When a concurrent new value
of P
load_min_set
and change in P
r_n
are commanded, the CM MAY wait to apply the change in P
r_n
at the next global
reconfiguration time (i.e., concurrent with the institution of the new value of P
load_min_set
) rather than applying the
change at the first sufficient reconfiguration time of the n
th
channel. The value of P
load_min_set
which applies to the new
P
r_n
is the concurrently commanded P
load_min_set
value, and thus if the change to P
r_n
falls outside the Dynamic Range
Window of the old P
load_min_set
, then the CM MUST wait for the global reconfiguration time to apply the change in
P
r_n
.
19

The CM MUST ignore the command to increase the per channel transmit power if such a command would cause P
load
n
for that channel to drop below P
load_min_set
. Note that the CMTS can allow small changes of power in the CM's
highest loaded channel, without these fluctuations impacting the transmit power dynamic range with each such small
change. This is accomplished by setting P
load_min_set
to a smaller value than normal, and fluctuation of the power per
channel in the highest loaded channel is expected to wander. The CM also MUST ignore any commanded change of
per channel transmit power which would result in P
r n
falling below the Dynamic Range Window, i.e., lower power
than permitted by the Dynamic Range Window, which is determined by P
load_min_set.
The CM also MUST ignore a
command changing P
load_min_set
such that existing values of P
r n
would fall outside the new Dynamic Range
Window.
20

16
Added equation per PHYv3.0-N-06.0340-1 by GO on 1/11/07.
17
Revised this sentence and following equation per PHYv3.0-N-06.0340-1 by GO on 1/11/07.
18
Revised this sentence per PHYv3.0-N-06.0340-1 by GO on 1/11/07.
19
Revised this paragraph and one Requirement per PHYv3.0-N-06.0341-1 by GO on 1/12/07.
20
68 CableLabs
11/17/11
The following two paragraphs define the CM and CMTS behavior in cases where there are Dynamic Range Window
violations due to indirect changes to P
hi_n,
or addition of a new channel with incompatible parameters without direct
change of P
r_n
or P
load_min_set
.
A change in the number of active channels changes P
hi_n
of the existing active channels (due to different P
hi_n
calculation tables for different number of active channels). Prior to changing the number of active channels, the
CMTS SHOULD change P
r_n
of all current active channels, if necessary, to fit in the new expected Dynamic Range
Window. If the change in the number of active channels would cause a violation of the Dynamic Range Window of
one of the existing active channels, the CM will ignore the change (i.e., not adding/removing an active channel). In
such case, the CM will also indicate partial service message in the REG-ACK or DBC response [MULPI].
When adding a new active channel to the transmit channel set, the new channel's power is calculated according to the
offset value defined in TLV 46.8.4 [MULPI], if it is provided. The CMTS SHOULD NOT set an offset value that
will result in a P
r_n
for the new channel outside the Dynamic Range Window. In the absence of the TLV, the new
channels power is initially set by the CM at the minimum allowable power, i.e., the bottom of the Dynamic Range
Window.
If one of the new active channels added by the change in the number of active channels violates the Dynamic Range
Window, the CM will ignore the change (i.e., not adding/removing an active channel). In such case, the CM will also
indicate partial service message in the REG-ACK or DBC response [MULPI].
For a UCD change, P
hi_n
may change due to a change in the modulation types. Additionally a UCD change that
modifies the Symbol Rate may result in a change in P
low_n
. The CMTS MUST NOT send a UCD change for the n
th

active channel that violates P
hi_n
P
load_min_set
P
low_n
. If a UCD changes the P
hi_n
for the n
th
active channel [MULPI],
the CM sets P
r_n
= P
hi_n
- P
load_n
for that channel, when the change count of the MAP matches the change count in the
new UCD. The new (changed) P
hi_n
and the current P
load_n
are used in the calculation. The CM maintains P
load_n
at its
value at the time the MAP change count matches the new UCD change count, and calculates a new P
r_n
and target
power for the channel, to be applied for bursts granted in the MAP with change count matching the change count in
the new UCD. The change in the transmit power P
r_n
due to the recalculation with the new P
hi_n
is not considered a
change in the commanded transmit power, for example as in Section 6.2.19, Section, 6.2.20, and Section 6.2.22,
and thus this change is required to be transparent on all active channels continuously bursting throughout such a
change. The CM is expected to meet all requirements without need of global reconfiguration time, even with the
induced change in P
r_n
, unless, of course, other parameters are changed which do imply the CM should be provided a
global reconfiguration time.
The spurious performance requirements of Sections 6.2.22.1, 6.2.22.1.1, and 6.2.22.1.2 apply when the CM is
operating within certain ranges of values for P
load_n
, for n = 1 to the active number of upstream channels, and for
certain ranges of the number of spreading codes being transmitted as a fraction of the total number of active codes on
a channel for S-CDMA channels, as detailed in those sections.
21

Transmit power per channel, for each channel, as reported by the CM in the MIB, is referenced to the 64-QAM
constellation. When transmitting with other constellations, a slightly different transmit power will result, depending
on the constellation gain in Table 67, Table 68, and Table 69 below. As an example, if the reported power in a
channel is 30 dBmV, 64-QAM will be transmitted with a target power of 30 dBmV in that channel, while QPSK will
be transmitted with 28.82 dBmV in that channel.

21
Paragraph modified per PHYv3.0-N-07.0599-1 on 1/25/08 by KN.
11/17/11 CableLabs
69
Table 67 - Constellation Gains and Per Channel Power Limits with
One Channel in the Transmit Channel Set
Constellation
Constellation
Gain Gconst
Relative to
64-QAM (dB)
Pmin
(dBmV)
Pmax
(dBmV)
TDMA
Pmax
(dBmV)
S-CDMA
Pmin - Gconst
(dBmV)
Pmax -
Gconst
(dBmV)
TDMA
Pmax - Gconst
(dBmV)
S-CDMA
L M H
QPSK -1.18 17 20 23 61 56 18.18 62.18 57.18
8-QAM -0.21 17 20 23 58 56 17.21 58.21 56.21
16-QAM -0.21 17 20 23 58 56 17.21 58.21 56.21
32-QAM 0.00 17 20 23 57 56 17.00 57.00 56.00
64-QAM 0.00 17 20 23 57 56 17.00 57.00 56.00
128-QAM 0.05 17 20 23 N/A 56 16.95 N/A 55.95
(Pmin is a function of Modulation Rate, with L = 1280 kHz, M = 2560 kHz, and H = 5120 kHz.)

Two Channels in the Transmit Channel Set
Constellation
Constellation
Gain Gconst
Relative to
64-QAM (dB)
Pmin
(dBmV)
Pmax
(dBmV)
TDMA
Pmax
(dBmV)
S-CDMA
Pmin - Gconst
(dBmV)
Pmax -
Gconst
(dBmV)
TDMA
Pmax - Gconst
(dBmV)
S-CDMA
L M H
QPSK -1.18 17 20 23 58 53 18.18 59.18 54.18
8-QAM -0.21 17 20 23 55 53 17.21 55.21 53.21
16-QAM -0.21 17 20 23 55 53 17.21 55.21 53.21
32-QAM 0.00 17 20 23 54 53 17.00 54.00 53.00
64-QAM 0.00 17 20 23 54 53 17.00 54.00 53.00
128-QAM 0.05 17 20 23 N/A 53 16.95 N/A 52.95

Three or Four Channels in the Transmit Channel Set
Constellation
Constellation
Gain Gconst
Relative to
64-QAM (dB)
Pmin
(dBmV)
Pmax
(dBmV)
TDMA
Pmax
(dBmV)
S-CDMA
Pmin -
Gconst
(dBmV)
Pmax -
Gconst
(dBmV)
TDMA
Pmax -
Gconst
(dBmV)
S-CDMA
L M H
QPSK -1.18 17 20 23 55 53 18.18 56.18 54.18
8-QAM -0.21 17 20 23 52 53 17.21 52.21 53.21
16-QAM -0.21 17 20 23 52 53 17.21 52.21 53.21
32-QAM 0.00 17 20 23 51 53 17.00 51.00 53.00
64-QAM 0.00 17 20 23 51 53 17.00 51.00 53.00
128-QAM 0.05 17 20 23 N/A 53 16.95 N/A 52.95

The actual transmitted power per channel, within a burst, MUST be constant to within 0.1 dB peak to peak, even in
the presence of power changes on other active channels. This excludes the amplitude variation, which is theoretically
present due to QAM amplitude modulation, pulse shaping, pre-equalization, and for S-CDMA, spreading and
varying number of allocated codes.
The CM MUST support the transmit power calculations defined in Section 6.2.19.1 and Section 6.2.19.2.

70 CableLabs
11/17/11
6.2.19.1 TDMA Transmit Power Calculations
In TDMA mode, the CM determines its target transmit power per channel P
t
as follows, for each channel that is
active. Define for each active channel, for example, upstream channel X:
P
r
= Reported power level (dBmV) of CM in MIB (refers to 64-QAM constellation) for channel X.
P = Power level adjustment (dB), for example, as commanded in ranging response message.
G
const
= Constellation gain (dB) relative to 64-QAM constellation (see above table).
P
min
= Minimum target transmit power per channel permitted for the CM per Section 6.2.19 (see above table).
P
max
= Maximum target transmit power permitted for the CM per Section 6.2.19 (see above table).
P
hi
= min(P
max
- G
const
) over all burst profiles used by the CM in channel X (see above table).
P
low
= max(P
min
- G
const
) over all burst profiles used by the CM in channel X (see above table).
P
t
= Target transmit power per channel level (dBmV) of CM in channel X (actual transmitted power per channel
in channel X as estimated by CM).
In the event of a UCD which changes the P
hi
for an active channel [MULPI], the CM sets P
r
= P
hi
- P
load
for that
channel when the MAP change count matches the change count in the new UCD, where the new (changed) P
hi
for the
channel and the current P
load
for the channel are used in the calculation.
22

The CM updates its reported power per channel in each channel by the following steps:
1. P
r
= P
r
+ P //Add power level adjustment (for each channel) to reported power level for each channel.
2. P
r
= min[P
r
, P
hi
] //Clip at max power limit per channel.
3. P
r
= max[P
r
, P
low
] //Clip at min power limit per channel.
4. Test, IF[P
r
 P
hi
- P
load
_
min
_
set
r
//Power per channel from this
command would violate the set Dynamic Range Window.
23

The CM then transmits, in channel X with target power per channel P
t
= P
r
+ G
const
, i.e., the reported power, plus the
constellation gain.
Usually the reported power level is a relatively constant quantity, while the transmitted power level in channel X
varies dynamically as different burst profiles, with different constellation gains, are transmitted. A CM's target
transmit power per channel MUST never be below P
min
or above P
max
. This implies that in some cases, the extreme
transmit power levels (e.g., 61 dBmV for QPSK and 17 dBmV) may not be permitted if burst profiles with multiple
constellations are active. Also, if only QPSK is used, the reported power per channel may be greater than 61 dBmV,
although the target transmit power per channel will not exceed 61 dBmV.
For example, if in channel X, only QPSK and 64-QAM burst profiles are active, P
hi
= 54 dBmV and P
low
= 18.2
dBmV for 1280 kHz modulation rate. P
low
is dependent on modulation rate. The maximum permitted QPSK
transmitted power in channel X is 54 dBmV - 1.2 dB= 52.8 dBmV, the minimum QPSK power in channel X is 18.2

22
Paragraph added per PHYv3.0-N008.0611-3 on 4/10/08 by KN.
23
Step modified per PHYv3.0-N-07-0465-1, #1 on 7/2/07 and PHYv3.0-N-07.0599-1 on 1/25/08 by KN.
11/17/11 CableLabs
71
dBmV - 1.2 dB= 17 dBmV (with 1280 kHz modulation rate), the maximum 64-QAM power in channel X is 54
dBmV, and the minimum 64-QAM power in channel X is 18.2 dBmV (with 1280 kHz modulation rate).
6.2.19.2 S-CDMA Transmit Power Calculations
In S-CDMA mode, the power calculations depend on whether the Maximum Scheduled Codes feature is enabled.
6.2.19.2.1 S-CDMA Transmit Power Calculations when Maximum Scheduled Codes Is Not Enabled
In S-CDMA mode when Maximum Scheduled Codes is not enabled, the CM determines its target transmit power per
channel P
t
, as follows, for each channel which is active. Define for each active channel, for example, upstream
channel X:
P
r
= reported power level (dBmV) of CM in MIB (refers to 64-QAM constellation and all active codes
transmitted) for channel X.
P
hi
= min[P
max
- G
const
] over all burst profiles used by the CM in channel X (see Table 67, Table 68, and Table
69).
P
low
= max[P
min
- G
const
] + 10 log (number_active_codes / number_of_codes_per_mini-slot) where the maximum
is over all burst profiles used by the CM in channel X (see Table 67, Table 68, and Table 69).
hi
r
= P
hi
- P
load
for that
hi
for the
load
24

25

1. P
r
= P
r
2. P
r
= min[P
r
, P
h
3. P
r
= max[P
r
, P
low
4. Test, IF[P
r
 P
hi
-P
load_min_set
r
In a spreader-on frame, the CM then transmits each code i with target power:
P
t,i
= P
r
+ G
const,i
- 10 log(number_active_codes)
i.e., the reported power for channel X, plus the constellation gain G
const,i
of that code, less a factor taking into account
the number of active codes. The total transmit power in channel X, P
t
, in a frame is the sum of the individual transmit
powers P
t,i
of each code in channel X, where the sum is performed using absolute power quantities [non-dB domain].
In a spreader-off frame, the CM target transmit power in channel X is P
t
= P
r
+ G
const
.
The transmitted power level in channel X varies dynamically as the number of allocated codes varies, and as
different burst profiles, with different constellation gains, are transmitted. A CM's target transmit power per channel
MUST never be below P
min
or above P
max
, including over all numbers of allocated codes and all burst profiles. This
implies that in some cases the extreme transmit power levels per channel (e.g., 17 and 56 dBmV) may not be

24
Paragraph added per PHYv3.0-N-08.0611-3 on 4/10/08 by KN.
25
Revised step 4 and 5 per PHYv3.0-N-06.0340-1 by GO on 1/11/07.
72 CableLabs
11/17/11
permitted. Also, if only QPSK is used, the reported power in a channel may be greater than 56 dBmV, although the
target transmit power per channel will not exceed 56 dBmV.
If, for example, the Transmit Channel Set contains only one channel, and this channel, channel X, has QPSK and 64-
QAM burst profiles, the number of active codes is 128 and the number of codes per mini-slot is 2, then P
hi
= 56
dBmV and P
low
= 36.24 dBmV for 1280 kHz modulation rate. P
low
is dependent on modulation rate. The maximum
permitted QPSK transmitted power in channel X is 56 1.18 = 54.82 dBmV when all active codes are transmitted.
The minimum QPSK power in channel X is 36.24 dBmV 1.18 dB 10log(128) dB + 10log(2) dB = 17 dBmV
(with 1280 kHz modulation rate) when one mini-slot is transmitted. The last term in the sum is the result of summing
the individual powers over two codes. Similarly, the maximum 64-QAM power in channel X is 56 dBmV, when all
active codes are transmitted and the minimum 64-QAM power in channel X is 36.24 dBmV 10log(128) dB +
10log(2) dB = 18.18 dBmV (with 1280 kHz modulation rate) when one mini-slot is transmitted. The minimum
QPSK power permitted in channel X while transmitting, for example, 2 mini-slots is 20 dBmV (with 1280 kHz
modulation rate), and the minimum 64-QAM power permitted in channel X, while transmitting 2 mini-slots, is 21.2
dBmV (with 1280 kHz modulation rate).
The CM needs to implement some form of clipping on the transmitted waveform at the higher output powers in order
to prevent peak to average ratio (PAR) issues.
The power received at the CMTS in a spreader-on frame will sometimes be less than the nominal power of a
spreader-off frame because of such factors as: 1) broadcast opportunities not used by any CM; 2) unicast grants not
used by one or more CMs; or 3) mini-slots assigned to the NULL SID.
6.2.19.2.2 S-CDMA Transmit Power Calculations When Maximum Scheduled Codes is Enabled
In S-CDMA mode on channels on which Maximum Scheduled Codes is enabled, the CM determines its target
transmit power per channel P
t
as follows, for each channel which is active. Define for each upstream channel, for
example, upstream channel X:
P
r
= reported power level (dBmV) of CM in MIB (operational transmit power of the spreader-off ranging burst
referenced to 64-QAM modulation) for channel X.
P
hi_S
= min[P
max
- G
const
] over all spreader-on burst profiles used by the CM in channel X (see Table 67, Table
68, and Table 69).
P
low_S
= max[17 - G
const
] + 10 log(number_active_codes / number_of_codes_per_mini-slot) where the maximum
is over all burst profiles used by the CM in channel X (see Table 67, Table 68, and Table 69).
P
max_T
= Maximum target transmit power permitted for the CM in channel X in TDMA mode (see Table 67,
Table 68, and Table 69) for the constellation used in ranging.
P
hi_T
= min[P
max_T
- G
const
] over all spreader-off burst profiles used by the CM in channel X (see Table 67,
Table 68, and Table 69).
P
on
= P
r
clipped at the maximum spreader-on limit.
P
sf
= CM Power Shortfall.
P
hr
= S-CDMA Power Headroom in dB. Equivalent to the value of the Ranging Response Message TLV-11
divided by 4.
P = power level adjustment in dB sent from CMTS to CM for channel X.
11/17/11 CableLabs
73
hi
r
= P
hi
- P
load
for that
hi
for the
load
26

The CM updates its power per channel in each channel by the following steps:
1. P
r
= P
r
2. P
r
= min[P
r
, P
hi_T
] //Clip at max TDMA power limit per channel.
3. P
r
= max[P
r
, P
low_S
] //Clip at min S-CDMA power limit per channel.
4. Test, IF[P
r
 P
hi_T
-P
load_min_set
r
6. P
on
= min[P
r
, P
hi_S
-P
load_min_set
] //Clip at max S-CDMA power limit per channel for the multiple channel dynamic
range limitation, as set by P
load_min_set
.
In spreader-off frames, the CM transmits in channel off with target power:
P
t
= P
r
+ G
const
Based on the spreader-off transmit power in channel X, the CM updates its power shortfall in channel X according to
the following steps:
with only one channel in the Transmit Channel Set:
P
sf
= P
t
- (56 - P
load_min_set
) // Difference between spreader-off and max spreader-on target powers in channel X;
P
sf
= max[P
sf
0] // Set P
sf
to 0 if P
t
is less than 56 dBmV - P
load_min_set
in channel X;
and with more than one channel in the Transmit Channel Set:
P
sf
= P
t
- (53 - P
load_min_set
) // Difference between spreader-off and max spreader-on target powers in channel X;
P
sf
= max[P
sf
, 0] // Set P
sf
to 0 if P
t
load_min_set
in channel X.
In spreader-on frames, the CM transmits each code i with target power:
P
t, i
= P
on
+ G
const, i
- 10 log(number_active_codes) + P
hr

i.e., the clipped reported power in channel X plus the constellation gain G
const, i
of that code, less a factor taking into
account the number of active codes, plus the Power Headroom P
hr
. P
hr
is the power (in dB) added to account for CMs
that have maximum scheduled code limits and can transmit additional power per code. The total transmit power in
channel X, P
t
, in a frame is the sum of the individual transmit powers P
t, i
of each code in channel X, where the sum
is performed over all N
alloc
allocated codes using absolute power quantities (non-dB domain).

26
74 CableLabs
11/17/11
If, for example, with more than one channel in the Transmit Channel Set, and the burst profile in channel X contains
QPSK for IUCs 1, 2, 3, and 4 and 64-QAM for IUCs 9 and 10, the number of active codes is 128, and the number of
codes per mini-slot is 2, then P
hi_S
= 53 dBmV, P
low_s
= 36.24 dBmV, and P
hi_T
= 58 dBmV. Assume P
load_min_set
= 0
dB. Assume the CM ranges channel X at spreader-off target transmit power of 57 dBmV. The CM reports for
channel X P
sf
= 57 dBmV - 53 dBmV = 4 dB. The CMTS uses P
sf
to set for channel X (using its vendor-specific
algorithm) max_scheduled_codes = 32 and P
hr
= 6 dB. (The S-CDMA power headroom may differ from the power
shortfall, at the discretion of the CMTS.) The CM sets its transmitted power per code in channel X to:
27

P
t, i
= P
on
+ G
const, i
- 10 log (number_active_codes) + P
hr
= 53 dBmV + 0 dB - 21 dB + 6 dB // For a code with 64-QAM modulation in channel X
= 38 dBmV
A parameter that may be used to illustrate the effect of increased power per code in channel X is the Effective
Transmit Power, P
eff
, the power that would result in channel X hypothetically if all N
a
active codes were transmitted.
It is computed as:

where the last term is the average constellation gain.
For a reference case with all codes transmitted in channel X using 64-QAM modulation (G
const
= 0 dB), the effective
transmit power reduces to:
P
eff
= P
on
+ P
hr
Continuing the above example, the result is:
P
eff
= 53 dBmV + 6 dB
= 59 dBmV
Limiting the number of codes has given the CM an enhanced effective power of 59 dBmV, which is 6 dB above the
normal maximum power per channel of 53 dBmV, and 2 dB above the ranging power of 57 dBmV. In this example,
the CMTS used its discretion to ask for 2 dB more enhancement in channel X than was needed (P
hr
= 6 dB vs P
sf
= 4
dB), perhaps due to some known impairment in the channel.
The effective_SNR is an SNR estimate for a given code corresponding to the effective transmit power in channel X.
It is defined as the measured SNR at the last station maintenance, minus the CM power shortfall, plus the power
headroom, plus the difference in constellation gain between the ranging burst and the code under consideration. Its
equation is:
effective_SNR = measured_SNR - P
sf
+ P
hr
+ (G
const, i
- G
const, ranging
)
where G
const
, ranging is the constellation gain of the ranging burst in channel X that resulted in the SNR
measurement.

27
11/17/11 CableLabs
75
In the MIB, effective_SNR corresponds to a reference case with 64-QAM modulation (G
const
,
i
= 0 dB):
sf
+ P
hr
- G
const, ranging
Continuing the example, if the measured SNR in channel X in the last station maintenance was 17 dB, using QPSK
modulation (G
const
,
ranging
= -1.2 dB), then the effective SNR referenced to 64-QAM modulation is:
effective_SNR = 17dB - 4dB + 6dB + 1.2 dB = 20.2 dB
6.2.19.3 Transmit Power Step Size
The step resolution in transmit power for each channel MUST be 0.5 dB or less. The step size accuracy MUST be
within 0.4 dB. For example, the actual power increase in a channel X resulting from a command to increase the
power level by 1 dB in a CM's next transmitted burst in channel X MUST be between 0.6 dB and 1.4 dB.
All active channels with no power change commanded MUST NOT change by more than 0.1 dB during a burst even
with a power change command to any subset of the other active channels.
After a change in the Dynamic Range Window (i.e., P
load_min_set
), the transmit power MUST NOT change by more
than 0.4 dB in any of the active channels when they resume transmission (referenced against any commanded change
in their transmit power).
A relaxation in step size accuracy to 1.4 dB is allowed for one gain change (and one setting of Dynamic Range
Window) for each 12 dB, when changing the power control range in either direction (from low-end to high-end
power and vice versa). The locations of these gain changes (increasing and decreasing to provide hysteresis), MUST
be at least 2 dB apart, thus enabling the use of large step attenuators in the coverage of the full power control range.
6.2.19.4 Transmit Power Requirements for Multiple Transmit Channel mode Not Enabled
28

With Multiple Transmit Channel mode disabled, the CM MUST operate as in [DOCSIS RFI 2.0], with the following
exceptions.
The minimum upstream power P
min
used by the CM MUST be as given in Table 610 below, superseding the
minimum upstream power requirements in [DOCSIS RFI 2.0].
The CM MUST replace the value of 53 in the equation for P
hi
_S in [DOCSIS RFI 2.0] with 56 since this value
is actually intended to be Pmax from Table 67 above.
Maximum upstream power prior to registration:
If the Extended Upstream Transmit Power Support TLV in the MDD is OFF, the CM MUST use a
maximum upstream power Pmax as given in Table 67 above, superseding the maximum upstream
power requirements in [DOCSIS RFI 2.0].
If the Extended Upstream Transmit Power Support TLV in the MDD is ON, the CM MAY exceed the
maximum upstream power Pmax as given in Table 67 above. The CM MUST NOT exceed a target
transmit power level of 61 dBmV. Note that all fidelity requirements specified in Section 6.2.22 still
apply when the CM is operating over an extended transmit power range.
Maximum upstream power after registration:
The CM MUST use a maximum upstream power Pmax given in Table 67 above, superseding the
maximum upstream power requirements in [DOCSIS RFI 2.0].

28
76 CableLabs
11/17/11
Table 610 - Minimum Transmit Power Pmin for Multiple Transmit Channel mode Not Enabled
Modulation Rate (kHz) Pmin (dBmV) Applicability

6.2.20 Burst Profiles
The transmission characteristics are separated into three portions: a) Channel Parameters; b) Burst Profile Attributes;
and c) User Unique Parameters. The Channel Parameters include: a) the modulation rate (six rates from 160 kHz to
5120 kHz in octave steps); b) the center frequency (Hz); c) the 1536-bit Preamble Superstring; and d) the S-CDMA
channel parameters. The Channel Parameters are shared by all users on a given channel. The Burst Profile Attributes
are listed in Table 611, and these parameters are the shared attributes corresponding to a burst type.
The CM MUST generate each burst at the appropriate time, as conveyed in the mini-slot grants provided by the
CMTS MAPs.
The CM MUST support all burst profiles commanded by the CMTS, via the Burst Descriptors in the UCD, and
subsequently assigned for transmission in a MAP.
Table 611 - Burst Profile Attributes
29

Burst Profile Attributes Configuration Settings
Modulation QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, 128-QAM(TCM Only)
Differential Encoding On/Off
TCM Encoding On/Off
Preamble Length 0-1536 bits (Note Section 6.2.10)
Preamble Value offset 0 to 1534
R-S FEC Error Correction (T) 0 to 16 (0 implies no R-S FEC. The number of codeword parity bytes is
2*T)
R-S FEC Codeword Information Bytes
(k)
Fixed: 16 to 253 (assuming R-S FEC on)
Shortened: 16 to 253 (assuming R-S FEC on)
Scrambler Seed 15 bits
Maximum Burst Size (mini-slots)
1
0 to 255
Guard Time Size 4 to 255 modulation intervals
There is no guard time in S-CDMA
Last Codeword Length Fixed, shortened
Scrambler On/Off On/Off
R-S Interleaver Depth (I
r
)
2
0 to floor (2048/N
r
)
3

29
Table modified per PHYv3.0-N-07.0465-1, # 2 on 7/2/07 and PHYv3.0-N-07.0598-1 on 1/25/08 by KN.
11/17/11 CableLabs
77
R-S Interleaver Block Size (B
r
)
4
2*N
r
to 2048
Preamble Type QPSK0/QPSK1
S-CDMA Spreader
5
On/Off
S-CDMA Codes per Subframe
5
1 to 128
S-CDMA Framer Interleaver Step Size
5
1 to (spreading intervals per frame - 1)
Table Notes:
Note 1 A max burst length of 0 mini-slots in the Burst Profile means that grant size is not constrained by the burst
profile parameter [MULPI].
Note 2 If depth=1, no interleaving; if depth=0, dynamic mode.
Note 3 Nr is the R-S code word size k+2T as defined in Section 6.2.5.1.
Note 4 Used only in dynamic mode
Note 5 Used only for S-CDMA channels.

The User Unique Parameters may vary for each user even when using the same burst type on the same channel as
another user (for example, Power Level), and are listed in Table 612.
Table 612 - User Unique Burst Parameters
User Unique Parameter Adjustment Command Resulting Parameter Value
Power Level
1
8-bit two's complement,
resolution = 0.25 dB
TDMA:
+17 to +57 dBmV (32-QAM, 64-QAM)
+17 to +58 dBmV (8-QAM, 16-QAM)
+17 to +61 dBmV (QPSK)
S-CDMA: +17 to +56 dBmV
(all modulations)
Resolution = 0.25 dB or better
Offset Frequency Range = 32 kHz, resolution = 1
Hz
Frequency Range per Section 6.2.17.1
Upstream Frequency Agility and Range
Ranging Offset Integer part: 32-bit two's
complement, resolution = (1 /
10.24 MHz) = 6.25 s/64 =
97.65625 ns

Fractional part: unsigned 8-bit
fractional extension, resolution =
6.25 s/(64*256) =
0.3814697265625 ns
Range: sufficient for maximum cable plant
length per Section 1.2.1

Resolution: Non-synchronous:
6.25s/64.
Synchronous: 6.25 s/(64*256)
30

Burst Length (mini-slots) if
variable on this channel
(changes burst-to-burst)
N/A 1 to 255 mini-slots

30
Column in row modified per PHYv3.0-N-06.0301-2, PO 11/20/06.
78 CableLabs
11/17/11
Transmit Equalizer Coefficients

(See Section 6.2.16, Transmit
Pre-Equalizer.)
DOCSIS 3.0 & 2.0 modes: 24
complex coefficients, 4 bytes per
coefficient
(2 real and 2 imaginary), load and
convolve modes

DOCSIS 1.1 mode: up to 64
coefficient (2 real and 2
imaginary), convolve mode only
DOCSIS 3.0 & 2.0 modes: 24 complex
coefficients

DOCSIS 1.1 mode: up to 64 complex
coefficients
Table Note:
Note 1 The minimum power level limit depends on modulation rate. The maximum power level limit depends on
modulation order, whether or not Multiple Transmit Channel mode is enabled and the number of upstream channels
in the Transmit Channel Set if MTC Mode is enabled.

The CM MUST implement the Offset Frequency Adjustment to effect a change in upstream carrier frequency within
10 Hz of the commanded change.
6.2.20.1 Ranging Offset
Ranging Offset is the time difference between the CM upstream frame time base and the CMTS upstream frame time
base. It is an advancement equal to roughly the round-trip delay between the CM and the CMTS, and is needed to
synchronize upstream transmissions in the TDMA and S-CDMA schemes. The CMTS MUST provide the CM with
feedback adjustments of this offset, based on reception of one or more successfully received bursts (i.e., satisfactory
result from each technique employed: error correction and/or CRC). The CMTS sends these Timing Adjust
commands to the CM in the Ranging Response MAC message, where a negative value implies the Ranging Offset is
to be decreased, resulting in later times of transmission at the CM.
The MDD message [MULPI] provides a TLV field that indicates whether the CMTS is providing a synchronous
downstream symbol clock, i.e., whether the downstream symbol clock is locked to the 10.24 MHz Master Clock. If
the MDD indicates that the CMTS is providing a synchronous downstream symbol clock, then the CM generates a
synchronous upstream symbol clock on all channels in the Transmit Channel Set, or in the single upstream channel if
MTC mode is disabled. If the MDD indicates that the CMTS is not providing a synchronous downstream symbol
clock, then the CM operates with a non-synchronous upstream for TDMA modulation and S-CDMA modulation is
not allowed. In the absence of an MDD message, the CM operates with a single upstream channel, which is a
synchronous upstream locked to the downstream symbol clock for S-CDMA modulation or a non-synchronous
upstream for TDMA modulation.
For non-synchronous channels the CM MUST implement the Timing Adjust command with resolution of at most 1
symbol duration (of the symbol rate in use for a given burst), and (other than a fixed bias) with accuracy within
0.25 s plus 1/2 symbol owing to resolution. As an example, for the maximum symbol rate of 5.12 Msym/s, the
corresponding symbol period would be 195 ns, the corresponding maximum resolution for the Timing Adjust MUST
be 195 ns. For the maximum symbol rate of 5.12 Msym/s, the corresponding minimum accuracy MUST be 348 ns.
The accuracy of CM burst timing of 0.25 s plus 1/2 symbol is relative to the mini-slot boundaries derivable at the
CM, based on an ideal processing of the timestamp signals received from the CMTS.
The resolution of the integer part of the Timing Adjust parameter, is (1 / 10.24 MHz) = 6.25 s/64 ~= 97.66 ns. For
S-CDMA channels, the CMTS provides an additional fractional field in the Timing Adjust command, with resolution
of 1/16384 of the frame tick increment = 6.25 s/(64*256) ~= 0.3814 ns. For S-CDMA channels, the CM MUST
implement the Timing Adjust to within 0.01 of the nominal chip period. As an example, for the maximum chip rate
of 5.12 MHz, the corresponding maximum resolution for implementation of the timing correction would be
(0.01)*195 ns or roughly 2 ns. For TDMA channels operating in synchronous mode, the CMTS is allowed to
11/17/11 CableLabs
79
provide the fractional field for the Timing Adjust. For TDMA channels operating in synchronous mode, when the
fractional field for the Timing Adjust is present in the RNG-RSP, the CM MUST implement the fractional field
Timing Adjust to within 0.01 of the nominal symbol period. For TDMA channels operating in synchronous mode,
if the RNG-RSP to a CM does not contain a fractional field Timing Adjust, the CM MUST assume that it is equal to
0 and still meet the accuracy requirement for implementing the fractional field Timing Adjust (to within 0.01 of the
nominal symbol period).
For CMTS/CM systems using multiple downstream channels, for a CM the CMTS designates a single channel as the
primary channel from which the CM derives CMTS master clock timing for all upstream transmission [MULPI].
31

For CMTS/CM systems using multiple upstream channels, the CM must range each upstream channel separately,
using the resolutions described above.
6.2.20.2 TDMA Reconfiguration Times
32

The CM MUST be capable of switching burst profiles with no reconfiguration time required between bursts, except
for changes in the following parameters: 1) Output Power; 2) Symbol Rate; 3) Offset Frequency; 4) Channel
Frequency; 5) Ranging Offset; and 6) Pre-Equalizer Coefficients. In the case of Multiple Transmit Channel mode
enabled, all of these parameters are adjusted independently on each upstream channel and reconfiguration time on
the non-adjusted channels is not required by the CM.
In the case of Multiple Transmit Channel mode enabled, global reconfiguration time (simultaneous on all of the
active channels) could be required by the CM when the Dynamic Range Window is adjusted (new value of
P
load_min_set
), or when any active channel accumulates a transmit power change of more than 3 dB (excluding transmit
power changes due to UCD-induced change in P
hi
[MULPI]) since the last global reconfiguration time (see Section
6.2.22.1). The CM MUST meet the requirements, on each active channel, for "Dynamic Range Window change"
(below) whenever P
load_min_set
is changed or re-commanded.
For Dynamic Range Window change or re-command: If P
load_min_set
is changed, or re-commanded, the CM MUST
be able to implement the change between bursts, as long as the CMTS allocates at least 96 symbols, plus 10 s,
between the last symbol center of one burst (on any active channel) and the first symbol center of the following burst
(on any active channel -- same channel or different channel from the previous burst). (Global "quiet" across all active
channels requires the intersection of ungranted burst intervals across all active TDMA channels to be at least 96
symbols plus 10 s duration.) The Output Power of the CM MUST be settled to within 0.1 dB of its final output
power level within 10 s from the beginning of the change in P
load_min_set
, and from the beginning of any Output
Power change on any channel that accompanies the change or re-command of P
load_min_set
. The Dynamic Range
Window, i.e., P
load_min_set
, MUST NOT be changed until the CM is provided sufficient time between bursts by the
CMTS. The Dynamic Range Window MUST NOT change while more than -30 dB of any symbol's energy of the
previous burst remains to be transmitted, or more than -30 dB of any symbol's energy of the next burst has been
transmitted.
For Output Power changes (excluding transmit power changes due to UCD-induced change in P
hi
when Multiple
Transmit Channel mode is enabled [MULPI]): If Output Power is to be changed by 1 dB or less, the CM MUST be
able to implement the change between bursts as long as the CMTS allocates at least 96 symbols, plus 5 s between
the last symbol center of one burst and the first symbol center of the following burst. If Output Power is to be
changed by more than 1 dB, the CM MUST be able to implement the change between bursts, as long as the CMTS
allocates at least 96 symbols, plus 10 s, between the last symbol center of one burst and the first symbol center of
the following burst. The maximum reconfiguration time of 96 symbols should compensate for the ramp-down time of
one burst and the ramp-up time of the next burst, as well as the overall transmitter delay time, including the pipeline
delay and pre-equalizer delay. The Output Power of the CM MUST be settled to within 0.1 dB of its final output
power level: a) within 5 s from the beginning of a change of 1 dB or less; and b) within 10 s from the beginning of
a change of greater than 1 dB. Output Power MUST NOT be changed until the CM is provided sufficient time

31
Section revised per PHYv3.0-N-06.0301-2 by PO 11/20/06.
32
Section revised per PHYv3.0-N-08.0611-3 on 4/10/08 by KN.
80 CableLabs
11/17/11
between bursts by the CMTS. Output Power MUST NOT change while more than -30 dB of any symbol's energy of
the previous burst remains to be transmitted, or more than -30 dB of any symbol's energy of the next burst has been
transmitted.
For Symbol Rate changes: The CM MUST be able to transmit consecutive bursts as long as the CMTS allows the
required time between bursts for UCD parameter changes. Symbol Rate MUST NOT be changed until the CM is
provided sufficient time between bursts by the CMTS. Symbol Rate MUST NOT change while more than -30 dB of
any symbol's energy of the previous burst remains to be transmitted, or more than -30 dB of any symbol's energy of
the next burst has been transmitted.
For Offset Frequency changes: The CM MUST be able to transmit consecutive bursts as long as the CMTS
allocates at least 96 symbols in between the last symbol center of one burst and the first symbol center of the
following burst. The maximum reconfiguration time of 96 symbols should compensate for the ramp-down time of
delay and optional pre-equalizer delay. Offset frequency MUST NOT be changed until the CM is provided sufficient
time between bursts by the CMTS. Offset frequency MUST NOT change while more than -30 dB of any symbol's
energy of the previous burst remains to be transmitted, or more than -30 dB of any symbol's energy of the next burst
has been transmitted.
For Channel Frequency changes: The CM MUST be able to implement the change between bursts as long as the
CMTS allocates at least 96 symbols plus 100 ms between the last symbol center of one burst and the first symbol of
the following burst. The Channel Frequency of the CM MUST be settled within the phase noise and accuracy
requirements of Section 6.2.22.5 and Section 6.2.22.6 within 100 ms from the beginning of the change. Channel
Frequency MUST NOT be changed until the CM is provided sufficient time between bursts by the CMTS. Channel
Frequency MUST NOT change while more than -30 dB of any symbol's energy of the previous burst remains to be
transmitted, or more than -30 dB of any symbol's energy of the next burst has been transmitted.
For Ranging Offset and/or Pre-Equalizer Coefficient changes: The CM MUST be able to transmit consecutive
bursts as long as the CMTS allocates at least 96 symbols in between the last symbol center of one burst and the first
symbol center of the following burst. The maximum reconfiguration time of 96 symbols should compensate for the
ramp-down time of one burst and the ramp-up time of the next burst, as well as the overall transmitter delay time,
including the pipeline delay and pre-equalizer delay. Ranging Offset MUST NOT be changed until the CM is
provided sufficient time between bursts by the CMTS. Ranging Offset MUST NOT change while more than -30 dB
of any symbol's energy of the previous burst remains to be transmitted, or more than -30 dB of any symbol's energy
of the next burst has been transmitted.
For Modulation Type changes: The CM MUST be able to transmit consecutive bursts with no reconfiguration time
between them (except for the minimum guard time). The modulation MUST NOT change while more than -30 dB of
the next burst has been transmitted, EXCLUDING the effect of the transmit equalizer (if present in the CM). (This is
to be verified with the transmit equalizer providing no filtering; delay only. Note that if the CMTS has decision
feedback in its equalizer, it may need to provide more than the 96 symbol gap between bursts of different modulation
type which the same CM may use; this is a CMTS decision.)
6.2.20.3 S-CDMA Reconfiguration Times
33

In S-CDMA mode, for changes in Output Power per mini-slot, Offset Frequency, Pre-equalizer coefficients, and/or
Ranging Offset, the CM MUST be able to transmit consecutive bursts as long as the CMTS allocates the time
duration of at least one frame in between the bursts. For all other burst profile parameter changes, no reconfiguration
is required beyond what is provided by the MAC for such changes. When Multiple Transmit Channel mode is

33
11/17/11 CableLabs
81
In the case of Multiple Transmit Channel mode enabled, global reconfiguration time could be required by the CM
when the Dynamic Range Window is adjusted (new value of P
load_min_set
), or when any active channel accumulates a
transmit power change of more than 3 dB (excluding transmit power changes due to UCD-induced change in P
hi

[MULPI]) since the last global reconfiguration time (see Section 6.2.22.1). (Global "quiet" across all active channels
requires the intersection of ungranted burst intervals across all active S-CDMA channels to be at least one frame.)
The CM MUST be able to transmit consecutive bursts, even with a change or re-command of P
load_min_set
, as long as
the CMTS allocates at least one frame in between bursts, across all channels in the Transmit Channel Set. (From the
end of a burst on one channel to the beginning of the next burst on any channel, there must be at least one frame
duration to provide a "global reconfiguration time" for S-CDMA channels.)
6.2.20.4 CM Timing Offsets When Changing Modulation Rate
When making a modulation rate change, the CM MUST employ the timing offsets shown in Table 613. The offsets
in the table correspond to the contribution of DOCSIS 1.0 and 1.1 legacy upstream receivers to changes in latency
when making modulation rate changes. The timing offset to apply is the difference between the entry in the table
corresponding to the new modulation rate and the entry corresponding to the original modulation rate. The offsets
are referenced to the center of the first symbol in the burst, which is the reference point for burst timing, as stated in
Section 6.2.21. Specification of these offsets is needed so that CMs apply uniform adjustments to their ranging
offsets, and CMTSs can appropriately handle CMs that apply these offsets when making modulation rate changes.
Table 613 - Timing Offset for Modulation Rate Changes
Modulation Rate (kHz) Timing Offset (in units of 1/64 time
ticks referenced to 5.12 MHz)
Applicability

For example, suppose a CM is on an upstream channel, operating at a modulation rate of 1.28 MHz. Now, suppose
the UCD message from the CMTS changes the modulation rate of the channel to 0.32 MHz. The CM applies an
additional timing offset of 168 24 = 144 to its ranging offset to compensate for this modulation rate change. The
value of 144 is positive, and thus, the CM will add to its ranging offset, so that it effectively transmits earlier by 144
units of 1/64 time ticks.
Furthermore, in changing modulation rates, if a CM has its own contribution to a change in latency, the CM MUST
also compensate for this CM-specific latency difference. This is in addition to the offset applied from the values in
the table above, which result from legacy CMTS upstream receiver contributions to changes in latency. The
requirements for CM burst timing accuracy found earlier in this section for TDMA mode, referenced to the
modulation rate that is the lower of the original, and the new modulation rate, apply after the modulation rate change,
with the required timing offsets above considered. Specifically, the CM MUST implement the timing adjustments
with accuracy within 0.25 s, plus 1/2 symbol, in both TDMA and S-CDMA modes.
A CMTS that does not apply the same internal physical delay offsets as the legacy DOCSIS upstream CMTS
receiver implementation is capable of receiving a CM burst after a modulation rate change in any of the following
ways, but is not limited to only these ways:
1. The CMTS may implement the internal physical delay offset, as specified in the above table.
2. The CMTS may implement an internal timing compensation based on the expected offset in the above table.
82 CableLabs
11/17/11
3. The CMTS may increase the guard time.
4. The CMTS may send an unsolicited RNG-RSP to each CM to adjust the delay offset. As discussed in [MULPI],
the CM is expected to be capable of adjusting its timing offset at any time with the accuracy specified within this
section.
6.2.21 Burst Timing Convention
Figure 631 illustrates the nominal burst timing for TDMA channels.
A) Nominal burst profile (no timing errors); 8
symbol guard band is illustrated; 10 symbol
ramp-up and ramp-down is illustrated.
B) Timing is referenced to the symbol
center of the first symbol of each burst.
Preamble
First
Symbol
Preamble
Last
Symbol
Data First
Symbol
Last FEC
Parity Symbol
Ramp down Ramp up
10
Symbols
8
Symbols
Burst N+1
Mini-slot
Boundary
Preamble
First Symbol
Center of Preamble
First Symbol = Mini-slot
Boundary
Burst N
Note: Ramp down of one burst can overlap ramp
up of following burst, even with on transmitter
assigned both bursts.

Figure 631 - Nominal TDMA Burst Timing

11/17/11 CableLabs
83
Figure 632 indicates worst-case burst timing for a TDMA channel. In this example, burst N arrives 1.5 symbols
late, and burst N+1 arrives 1.5 symbols early, but separation of 5 symbols is maintained; the 8-symbol guard band is
shown below:
Burst N+1
Mini-slot
Boundary
Burst N
Worst case Burst timing
input to CMTS
Time equal to 5 symbols, separates the first
symbol center of burst N+1 and the last
symbol center of burst N
Symbol center of
Burst N+1, last symbol;
1.5 symbols early
Symbol center of
Burst N, last symbol;
1.5 symbols early
8 symbol
guard band
5
symbols
1.5 1.5

Figure 632 - Worst-Case TDMA Burst Timing

At a symbol rate of R
s
, symbols occur at a rate of one each T
s
= 1/R
s
s. Ramp-Up and Ramp-Down are the spread of
a symbol in the time domain beyond T
s
duration, owing to the symbol-shaping filter, and any residual effect from the
transmit equalizer. If only one symbol were transmitted, its duration would be longer than T
s
due to the shaping filter
impulse response being longer than T
s
. The spread of the first and last symbols of a burst transmission effectively
extends the duration of the burst to longer than N * T
s
, where N is the number of symbols in the burst.
For S-CDMA channels, the bursts from all CMs are synchronized. This means that the ramp-down of one burst may
occur at the same time as the ramp-up of the subsequent burst. The CM MUST meet the ranging and synchronization
requirements of S-CDMA to assure that the ramp-down and ramp-up of bursts are aligned.
6.2.22 Fidelity Requirements
The following requirements assume that any pre-equalization is disabled, unless otherwise noted.
6.2.22.1 Spurious Emissions
The noise and spurious power MUST NOT exceed the levels given in Table 614, Table 615, and Table 617. The
noise and spurious power SHOULD NOT exceed the levels given in Table 616 and Table 618. When Multiple
Transmit Channel mode is enabled, the noise and spurious power requirements for the individual channels are
summed (absolute power, NOT in dB) to determine the composite noise floor for the multi-channel transmission
condition. Also when Multiple Transmit Channel mode is enabled, these spurious performance requirements only
apply when the CM is operating within certain ranges of values for P
load_n
, for n = 1 to the number of upstream
channels in the Transmit Channel Set, and for the ratio of number_allocated_codes / number_active_codes from 1
down to 1/4
th
, for S-CDMA channels.
84 CableLabs
11/17/11
For S-CDMA mode, when a modem is transmitting fewer than 1/4th of the active spreading codes on a channel, the
spurious emissions requirement limit is the power value (in dBmV), corresponding to the specifications for the
power level associated with 1/4
th
of the active spreading codes (i.e., 6 dB below the channel power corresponding to
all active codes for that channel allocated to the CM).
When Multiple Transmit Channel mode is enabled and there are two or more channels in the Transmit Channel Set,
the spurious performance requirements MUST be met only when the following loading ranges are satisfied:
Two or more channels in the TCS: applicable when P
load_1
P
load_min_set
3 dB, AND
Two channels in the TCS: applicable when P
load_2
P
load_1
12 dB.
Three channels in the TCS: applicable when P
load_3
P
load_1
8 dB.
Four channels in the TCS: applicable when:
P
load_2
P
load_1
4 dB.
AND P
load_3
P
load_2
4 dB.
34

AND P
load_4
P
load_3
4 dB.
When a modem is transmitting at power levels falling outside these loading ranges (which is allowed by the 12 dB
dynamic range), the spurious emissions requirement limits are the power values (in dBmV rather than dBc)
corresponding to the specifications associated with transmissions at the second and higher loaded channels if the
loading of those channels were decreased (transmit power increased) to just meet the applicability ranges.
With Multiple Transmit Channel mode enabled, the spurious performance requirements do not apply to any upstream
channel from the time the output power on any active upstream channel has varied by more than 3 dB since the last
global reconfiguration time through the end of the next global reconfiguration time changes, excluding transmit
hi
[MULPI]. "Global reconfiguration time" is defined as the inactive
time interval provided between active transmissions, which simultaneously satisfies the requirement in Section
6.2.20.2 for all TDMA channels in the TCS and the requirement in Section 6.2.20.3 for all S-CDMA channels in the
TCS. With Multiple Transmit Channel mode enabled, the CMTS SHOULD provide global reconfiguration time to a
CM before (or concurrently as) the CM has been commanded to change any upstream channel transmit power by 3
dB cumulative since its last global reconfiguration time, excluding transmit power changes due to UCD-induced
change in P
hi
[MULPI].
35

As one example, with three S-CDMA channels in the TCS, at transmit powers per channel of 53 dBmV, 41 dBmV,
and 41 dBmV, the absolute spurious emissions requirements, when two or more channels in the TCS are
transmitting, correspond to the absolute spurious emissions requirements with transmit power per channel of 53
dBmV, 45 dBmV, and 45 dBmV. The 41 dBmV power levels are raised for the purposes of determining the spurious
emissions limits, up to 53 dBmV 8 dB = 45 dBmV.
As an example, with four S-CDMA channels in the TCS, at transmit powers per channel of 53 dBmV, 41 dBmV, 41
dBmV, and 41 dBmV, the absolute spurious emissions requirements, when two or more channels in the TCS are
dBmV, 49 dBmV, 45 dBmV, and 41 dBmV. The 41 dBmV power levels are raised for the purposes of determining
the spurious emissions limits, up to 53 dBmV 4 dB = 49 dBmV; and the artificial 49 dBmV 4 dB = 45 dBmV;
and the artificial 45 dBmV 4 dB = 41 dBmV.

34
Paragraph revised per PHYv3.0-N-06.0340-1 by GO on 1/11/07.
35
Paragraph revised per PHYv3.0-N-08.0611-3 on 4/10/08 by KN.
11/17/11 CableLabs
85
In Table 614, inband spurious emissions includes noise, carrier leakage, clock lines, synthesizer spurious products,
and other undesired transmitter products. It does not include ISI. The measurement bandwidth for Inband spurious is
equal to the modulation rate (e.g., 1280 to 5120 kHz). All requirements expressed in dBc are relative to the actual
transmit power that the CM emits in one channel.
The measurement bandwidth is 160 kHz for the Between Bursts (none of the channels in the TCS are bursting) specs
of Table 614, except where called out as 4 MHz or 250 kHz.
The Transmitting Burst specs apply during the mini-slots granted to the CM (when the CM uses all or a portion of
the grant), and for 32 modulation intervals before and after the granted mini-slots. The Between Bursts specs apply
except during a used grant of mini-slots on any active channel for the CM, and the 32 modulation intervals before
and after the used grant.
In TDMA mode, a mini-slot may be as short as 32 modulation intervals, or 6.25 s at the 5.12 MHz rate, or as short
as 25 s at the 1.28 MHz rate.
36

Table 614 - Spurious Emissions
37

Parameter Transmitting Burst Between Bursts
Inband -40 dBc -72 dBc
Adjacent Band See Table 615 and
Table 616.
-72 dBc
Within the upstream operating range 5-42
MHz or 5-85 MHz (excluding assigned
channel, adjacent channels)
See Table 617 and
Table 618.
-72 dBc
For the case where the upstream operating
range is 5-42 MHz:
CM Integrated Spurious Emissions Limits
(all in 4 MHz, includes discretes)
1
42 to 54 MHz
54 to 60 MHz
60 to 88 MHz
88 to 870 MHz, or 1002 MHz if CM center
frequency ranges up to 999 MHz

-40 dBc
-35 dBmV
-40 dBmV
-45 dBmV

-26 dBmV
-40 dBmV
-40 dBmV
max(-45 dBmV, -40 dB ref d/s
2
)
range is 5-42 MHz:
CM Discrete Spurious Emissions Limits
1
42 to 54 MHz
54 to 88 MHz

-50 dBc
-50 dBmV
-50 dBmV

-36 dBmV
-50 dBmV
-50 dBmV

36
Paragraph modified per PHYv3.0-N-07.0465-1 #3 on 7/2/07 by KN.
37
Table modified per PHYv3.0-N-07.0456-1 on 6/12/07, per PHYv3.0-N-07.0465-1 on 7/2/07, and per PHYv3.0-N-08.0361-1
on 4/17/08 by KN.
86 CableLabs
11/17/11
range is 5-85 MHz:
CM Integrated Spurious Emissions Limits
(all in 4 MHz, includes discretes)
3
85 to 108 MHz
85 to 108 MHz (SHOULD)
108 to 136 MHz

-45 dBc
-50 dBc
-40 dBmV
-45 dBmV

-31 dBmV
-36 dBmV
-40 dBmV
4
)
range is 5-85 MHz:
1
85 to 108 MHz

-50 dBc
-50 dBmV

-36 dBmV
-50 dBmV
Table Notes:
Note 1 These spec limits exclude a single discrete spur related to the tuned received channel; this single
discrete spur MUST be no greater than -40 dBmV.
Note 2 "dB ref d/s" is relative to the received downstream signal level. Some spurious outputs are proportional to
the receive signal level.
Note 3 These spec limits exclude a single discrete spur related to the tuned received channel; this spur MUST be
no greater than -40 dBmV.
Note 4 "dB ref d/s" is relative to the received downstream signal level.

6.2.22.1.1 Adjacent Channel Spurious Emissions
Spurious emissions from a transmitted carrier may occur in an adjacent channel which could be occupied by a carrier
of the same or different modulation rate. Table 615 lists the required adjacent channel spurious emission levels for
all combinations of transmitted carrier modulation rates and adjacent channel modulation rates. The measurement is
performed in an adjacent channel interval of appropriate bandwidth and distance from the transmitted carrier, based
on the modulation rates of the transmitted carrier, and the carrier in the adjacent channel.
Table 615 - Adjacent Channel Spurious Emissions Requirements Relative to the Per Channel Transmitted
Burst Power Level for Each Channel
Specification in the
interval
Measurement interval and distance
from carrier edge
Adjacent channel carrier
modulation rate
-50 dBc 20 kHz to 180 kHz 160 kHz
-50 dBc 160 kHz to 1440 kHz 1280 kHz
-47 dBc 320 kHz to 2880 kHz 2560 kHz
-44 dBc 640 kHz to 5760 kHz 5120 kHz

11/17/11 CableLabs
87
In addition, the CM SHOULD meet the requirements in Table 616, under the same conditions as listed for Table 6
15 above. With a lower power channel bursting as a lone bursting channel (at that point in time), 2 dB relaxation will
apply to the absolute spurious emissions requirement of Table 616.
Table 616 - Adjacent Channel Spurious Emissions Goals Relative to the Per Channel Transmitted Burst
Power Level for Each Channel
interval
Measurement interval and distance
from carrier edge
modulation rate
-51 dBc 160 kHz to 1440 kHz 1280 kHz
-50 dBc 320 kHz to 2880 kHz 2560 kHz
-48 dBc 640 kHz to 5760 kHz 5120 kHz

6.2.22.1.2 Spurious Emissions in the Upstream Frequency Range
Table 617 lists the possible modulation rates that could be transmitted in an interval, the required spurious level in
that interval, and the initial measurement interval at which to start measuring the spurious emissions. Measurements
should start at the initial distance and be repeated at increasing distance from the carrier until the upstream band edge
is reached. The lower band edge of the upstream is 5 MHz; the upper band edge of the upstream is 42 MHz or if
upstream frequency range selection is available in the CM, 85 MHz may be selected.
Table 617 - Spurious Emissions Requirements in the Upstream Frequency Range Relative to the Per
Channel Transmitted Burst Power Level for Each Channel
Possible modulation rate in
this interval
interval
Initial measurement interval and
distance from carrier edge
160 kHz -54 dBc 220 kHz to 380 kHz
1280 kHz -50 dBc 360 kHz to 1640 kHz
2560 kHz -47 dBc 520 kHz to 3080 kHz
5120 kHz -44 dBc 840 kHz to 5960 kHz

In addition, the CM SHOULD meet the requirements in Table 618, under the same conditions as listed for Table 6
17 above. With a lower power channel bursting as a lone bursting channel (at that point in time), 2 dB relaxation will
apply to the absolute spurious emissions requirement of Table 618.
Table 618 - Spurious Emissions Goals in the Upstream Frequency Range Relative to the Per Channel
Transmitted Burst Power Level for Each Channel
this interval
interval
88 CableLabs
11/17/11
this interval
interval
1280 kHz -54 dBc 360 kHz to 1640 kHz
2560 kHz -52 dBc 520 kHz to 3080 kHz
5120 kHz -50 dBc 840 kHz to 5960 kHz

6.2.22.2 Spurious Emissions During Burst On/Off Transients
Each transmitter MUST control spurious emissions, prior to and during ramp-up, during and following ramp-down,
and before and after a burst.
On/off spurious emissions, such as the change in voltage at the upstream transmitter output, due to enabling or
disabling transmission, MUST be no more than 100 mV. Such a step MUST be dissipated no faster than 2 s of
constant slewing. This requirement applies when the CM is transmitting at +55 dBmV or more per channel on any
channel. At backed-off transmit levels, the maximum change in voltage MUST decrease by a factor of 2 for each 6
dB decrease of power level in the highest power active channel, from +55 dBmV per channel, down to a maximum
change of 7 mV at 31 dBmV per channel and below. This requirement does not apply to CM power-on and power-
off transients.
38

6.2.22.3 Modulation Error Ratio (MER)
MER measures the cluster variance caused by the transmit waveform. It includes the effects of ISI, spurious, phase
noise, and all other transmitter degradations.
6.2.22.3.1 Definitions
Symbol MER: MER
symb
is defined as follows for TDMA or S-CDMA symbols. The transmitted RF waveform (after
appropriate down conversion) is applied to the ideal receive symbol matched filter and is sampled once per symbol.
For TDMA, the matched filter is a square-root raised cosine filter with alpha = 0.25. For S-CDMA, the matched
filter is a square-root raised cosine filter with alpha = 0.25, convolved with the time-reversed spreading code
sequence. (In this convolution, the spreading code sequence is expressed as a weighted impulse train spaced at the
chip period.) No external noise (AWGN) is added to the signal. The carrier frequency offset, carrier phase offset,
symbol timing and gain may be adjusted during each burst to maximize MER
symb
. Equalization of the received
waveform is not permitted. For cases where the CM transmit equalizer is ON, the transmit equalizer coefficients may
be adjusted to maximize MER
symb
. MER
symb
is defined at the F connector of the CM, except that when an echo
channel is inserted, MER
symb
is defined at the output of the echo channel. MER
symb
is computed by the formula:
( )
=
N
j
j
av
symb
e
N
E
dB MER
1
2
10
1
log 10

where:
E
av
is the average constellation energy for equally likely symbols (see Section 6.2.14 and Figure 620),
N is the number of symbols averaged;

38
Revised paragraph per PHYv3.0-N-06.0341-1 by GO on 1/12/07.
11/17/11 CableLabs
89
ej is the error vector from the jth received symbol to the ideal transmitted QAM symbol on the grid of
Figure 620.
For S-CDMA, MER
symb
is averaged over all active codes.
MER of composite chips: MER
chip
is specified for composite S-CDMA chips to ensure that high SNR is
maintained, especially for a small number of allocated codes, to prevent noise funneling effects when many modems
transmit simultaneously. A composite S-CDMA chip is defined as the output of the spreader during one chip
interval, that is, an element of the transmission vector defined in Section 6.2.15, "S-CDMA Spreader".
MER
chip
is defined as follows: The transmitted RF waveform (after appropriate down conversion) is applied to the
ideal receive chip matched filter and is sampled once per chip. The matched filter is a square-root raised cosine filter
with alpha = 0.25. No external noise (AWGN) is added to the signal. The carrier frequency offset, carrier phase
offset, timing and gain may be adjusted during each burst to maximize MER
chip
. Equalization of the received
waveform is not permitted. For cases where the CM transmit equalizer is ON, the transmit equalizer coefficients may
be adjusted to maximize MER
chip
. MER
chip
is defined at the F connector of the CM. MER
chip
is computed by the
formula where:
( )
=
=
N
j
j j
N
j
j
chip
r p
p
dB MER
1
2
1
2
10
log 10
where:
p
j
is the jth ideal transmitted composite chip
r
j
is the jth received composite chip
N is the number of composite chips observed
6.2.22.3.2 Requirements
Unless otherwise stated, the MER MUST meet or exceed the following limits over the full transmit power range of
Table 612 for each modulation, modulation rate, and over the full carrier frequency range, and for S-CDMA, over
any valid number of active and allocated codes. The 5-85 MHz carrier frequency range refers more precisely to [5
MHz + modulation rate * 1.25 / 2] to [85 MHz - modulation rate * 1.25 / 2]. At the break points between regions, the
higher MER specification applies.
Case 1: Flat channel, transmit equalization OFF
Case 1a: for modulation rates 2.56 MHz and below for 5 MHz to 42 MHz mode of operation:
MER
symb
30 dB over 15 to 30 MHz carrier frequency
MER
symb
27 dB over 10 MHz to 15 MHz and 30 MHz to 35 MHz carrier frequency
MER
symb
Case 1b: for modulation rate 5.12 MHz for 5 MHz to 42 MHz mode of operation:
MER
symb
MER
symb
MER
symb
90 CableLabs
11/17/11
Case 1c: for modulation rates 2.56 MHz and below for 5 MHz to 85 MHz mode of operation:
MER
symb
MER
symb
MER
symb
Case 1d: for modulation rate 5.12 MHz for 5 MHz to 85 MHz mode of operation:
MER
symb
MER
symb
MER
symb
Case 2: Flat channel, transmit equalization ON
Case 2a: for TDMA/QPSK, MER
symb
30 dB.
Case 2b: for S-CDMA and all TDMA modulations except QPSK, MER
symb
35 dB.
Case 2c: for S-CDMA, MER
chip
33 dB.
Case 3: Echo channel, transmit equalization ON
39

(Echo channel MUST NOT be limited by dynamic range. Echo channel MUST be a true representation described in
Table 52)
Case 3a: In the presence of a single echo selected from the channel micro-reflections defined in Table 52, the
measured MERsymb MUST be 30 dB for TDMA/QPSK, and 33 dB for S-CDMA and all TDMA modulations,
except QPSK.
Case 3b: In the presence of two or three of the echoes defined in Table 52 (at most one of each specified magnitude
and delay), the measured MER
symb
MUST be 29 dB.
Since the table does not bound echo delay for the -30 dBc case, for testing purposes, it is assumed that the time span
of the echo at this magnitude is less than or equal to 1.5 s.
The CMTS MUST provide a test mode in which it:
accepts equalizer coefficients via an external interface, (e.g., Ethernet).
sends the coefficients to the CM's pre-equalizer via ranging response message (both load and convolve modes).
does not adjust the CM's frequency, timing or power.
6.2.22.4 Filter Distortion
The following requirements assume that any pre-equalization is disabled.
6.2.22.4.1 Amplitude
The spectral mask MUST be the ideal square-root raised-cosine spectrum with alpha = 0.25, within the ranges given
in Table 619.

39
Revised Case 3 per PHYv3.0-N-06.0341-1 by GO on 1/12/07.
11/17/11 CableLabs
91
Table 619 - Single Channel Filter Amplitude Distortion
Frequency Spectral Mask Limits
Lower Upper
f
c
- 5R
s
/8 - -30dB
f
c
- R
s
/2 -3.5dB -2.5dB
f
c
- 3R
s
/8 to f
c
- R
s
/4 -0.5dB +0.3dB
f
c
- R
s
/4 to f
c
+ R
s
/4 -0.3dB +0.3dB
f
c
+ R
s
/4 to f
c
+ 3R
s
/8 -0.5dB +0.3dB
f
c
+ R
s
/2 -3.5dB -2.5dB
f
c
+ 5R
s
/8 - -30dB

Where f
c
is the center frequency, R
s
is the modulation rate, and the spectral density is measured with a resolution
bandwidth of 10 kHz or less.
6.2.22.5 Carrier Phase Noise
The upstream transmitter total integrated phase noise (including discrete spurious noise) MUST be less than or equal
to -46 dBc, summed over the spectral regions spanning 200 Hz to 400 kHz above and below the carrier.
The upstream transmitter total integrated phase noise (including discrete spurious noise) MUST be less than or equal
to -44 dBc, summed over the spectral regions spanning 8 kHz to 3.2 MHz above and below the carrier.
The CM MUST provide a test mode in which:
A continuous (non-bursted), unmodulated (CW) upstream signal is transmitted at the commanded carrier
frequency, modulation rate and level. This is equivalent to replacing the chip sequence at the spreader output
with the constant sequence (1, 1, 1, 1, 1, 1,...) at nominal amplitude, equal on both I and Q.
The CM tracks the downstream symbol clock and uses it to generate the upstream symbol clock as in normal
synchronous operation.
6.2.22.6 Channel Frequency Accuracy
The CM MUST implement the assigned channel frequency within 50 parts per million, over a temperature range of 0
to 40 degrees C, and up to five years from date of manufacture.
6.2.22.7 Modulation Rate Accuracy
The MDD message [MULPI] provides a TLV field that indicates whether the CMTS is providing a synchronous
downstream symbol clock, i.e., whether the downstream symbol clock is locked to the 10.24 MHz Master Clock. If
the MDD indicates that the downstream symbol clock is synchronous, the CM MUST lock its timing to the
downstream symbol clock. If the MDD indicates that the downstream symbol clock is asynchronous, the CM MUST
NOT lock its timing to the downstream symbol clock.
In synchronous operation (i.e., when the CM is locked to the downstream symbol clock), the upstream modulator
MUST lock the upstream modulation rate of all upstream channels (S-CDMA and/or TDMA) to the downstream
symbol rate, subject to the symbol timing jitter requirements of Section 6.2.22.8.
92 CableLabs
11/17/11
In asynchronous operation (i.e., when the CM timing is not locked to the downstream symbol clock), the upstream
modulator MUST provide an absolute modulation rate accuracy of symbol rates, 50 parts per million, over a
temperature range of 0 to 40 degrees C, up to five years from date of manufacture.
6.2.22.8 Modulation Timing Jitter
6.2.22.8.1 Symbol Timing Jitter for TDMA Upstream Channels in Asynchronous Operation
For TDMA upstream channels in asynchronous operation, peak-to-peak symbol jitter, referenced to the previous
symbol zero-crossing of the transmitted waveform, MUST be less than 0.02 of the nominal symbol duration over a 2
s period. In other words, the difference between the maximum and the minimum symbol duration during the 2 s
period shall be less than 0.02 of the nominal symbol duration for each of the five upstream symbol rates.
For TDMA upstream channels in asynchronous operation, the peak-to-peak cumulative phase error, referenced to the
first symbol time, and with any fixed symbol frequency offset factored out, MUST be less than 0.04 of the nominal
symbol duration over a 0.1 s period. In other words, the difference between the maximum and the minimum
cumulative phase error during the 0.1 s period shall be less than 0.04 of the nominal symbol duration for each of the
five upstream symbol rates. Factoring out a fixed symbol frequency offset is to be done by using the computed mean
symbol duration during the 0.1 s.
6.2.22.8.2 Modulation Timing Jitter for Synchronous Operation
All jitter specifications assume a downstream input to the CM per Sections 6.3.5, 6.3.3 and [DRFI].
For S-CDMA and TDMA upstream channels in synchronous operation, upstream modulation clock timing error
(with the mean error subtracted out) relative to the CMTS master clock MUST be less than 0.005 RMS of the
modulation interval over a 35 s measurement duration. This applies to: 1) the worst-case jitter and frequency drift
specified for the CMTS Master clock and the CMTS downstream symbol clock in the requirements above; and 2) for
any round-trip propagation delay up to the maximum allowed.
The CM upstream modulation clock SHOULD track the jitter components below 10 Hz in the input downstream
symbol clock with an error transfer function below -25 dB. The CM upstream modulation clock SHOULD attenuate
the jitter components in the input downstream symbol clock above 200 Hz.
The CM MUST provide a test mode in which:
A continuous (non-bursted) upstream signal is transmitted at the commanded carrier frequency, modulation rate
and level.
The modulation sequence at the transmit equalizer input is replaced with an alternating binary sequence (1, -1, 1,
-1, 1, -1,...) at nominal amplitude, equal on both I and Q.
The CM tracks the downstream symbol clock and uses it to generate the upstream modulation clock as in normal
synchronous operation.
6.2.23 Upstream Demodulator Input Power Characteristics
The instantaneous input signal level, including ingress and noise to the upstream demodulator, MUST NOT exceed
29 dBmV in the 5-85 MHz frequency range of operation. The intended received power in each carrier MUST be
within the values shown in Table 620.
The demodulator MUST operate within its defined performance specifications with received bursts within 6 dB of
the nominal commanded received power. Thus the maximum power in a received burst MUST be less than or equal
29 dBmV.
11/17/11 CableLabs
93
Table 620 - Upstream Channel Demodulator Input Power Characteristics
Modulation Rate (kHz) Maximum Range (dBmV) Applicability
160 -13 to +17 CMTS MAY support
1280 -13 to +17 CMTS MUST support

6.2.24 Upstream Electrical Output from the CM
The CM MUST output an RF modulated signal with the characteristics delineated in Table 621.
Table 621 - Electrical Output from CM
40

Parameter Value
Frequency 5 to 42 MHz edge to edge
The following option MAY be provided:
Mode Selection Option:
Mode 1: 5 to 42 MHz edge to edge
Level range per channel (Multiple
Transmit Channel mode disabled,
or only Multiple Transmit Channel
mode enabled with one channel in
the TCS)
TDMA:
P
min
to +57 dBmV (32-QAM, 64-QAM)
P
min
P
min
to +61 dBmV (QPSK)
S-CDMA:
P
min
to +56 dBmV (all modulations)
where
P
min
= +17 dBmV, 1280 kHz modulation rate
P
min
P
min
Level range per channel
(two channels in the TCS)
TDMA:
P
min
P
min
P
min
to +58 dBmV (QPSK)
S-CDMA:
P
min
where
P
min
P
min
P
min

40
Table modified per PHYv3.0-N-07.0456-1 #3 on 6/12/07 by KN.
94 CableLabs
11/17/11
Parameter Value
(three or four channels in the TCS)
TDMA:
P
min
P
min
P
min
to +55 dBmV (QPSK)
S-CDMA:
P
min
where
P
min
P
min
P
min
Modulation Type QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, and 128-QAM
Modulation Rate (nominal) TDMA: 1280, 2560, and 5120 kHz
S-CDMA: 1280, 2560, and 5120 kHz
Optional pre-3.0-DOCSIS operation,
TDMA: 160, 320, and 640 kHz
Bandwidth TDMA: 1600, 3200, and 6400 kHz
S-CDMA: 1600, 3200, and 6400 kHz
Optional pre-3.0-DOCSIS operation,
TDMA: 200, 400, and 800 kHz
Output impedance 75 ohms
Output Return Loss >6 dB (across the selected upstream frequency range 5-42 MHz or 5-85 MHz).
> 6 dB (108 to 870 MHz, or 108 to 1002 MHz if CM center frequency ranges
up to 999 MHz)
Connector F connector per [ISO/IEC-61169-24] or [SCTE 02] (common with the input)

6.2.25 Upstream CM Transmitter Capabilities
The CM advertises its capabilities to the CMTS. These capabilities include:
Maximum number of active channels, which MUST be 4 or greater.
Maximum number of 3.2 MHz channels, which MUST be no greater than the maximum number of active
channels supported.
Maximum number of 6.4 MHz channels, which MUST be no greater than the maximum number of 3.2 MHz
channels supported. The maximum number of 6.4 MHz channels MUST be 4 or greater.
41

Selectable Active Codes Mode 2 capability yes/no.
42

Code Hopping Mode 2 capability yes/no.
Supports Extended Upstream Frequency Range (5 - 85 MHz) yes/no.
Extended Upstream Transmit Power capability yes/no plus capability in range of 51.25 dBmV to 61 dBmV.
43

41
42
Text revised per PHYv3.0-N-07.0465-1, # 5 on 7/2/07 by KN.
43
11/17/11 CableLabs
95
The CM MUST support all the capability it reports to the CMTS.
The CMTS MUST NOT command the CM to operate outside its advertised capabilities.
6.2.25.1 Description of CM Upstream Transmit Channel Set Capability
The CM communicates to the CMTS its capabilities in regards to the number of active upstream channels it can
support. The method of conveyance must clearly and unambiguously describe the capabilities of the CM to the
CMTS. The parameters for this messaging are defined in this section, and Appendix IV contains example
calculations for this reporting from the CM to the CMTS of the number of active upstream channels of the various
bandwidths supported by the CM.
Let X = number of 6.4 MHz channels, Y = number of 3.2 MHz channels, and Z = number of 1.6 MHz channels that
are active at one time (the Transmit Channel Set). Let X
max
be the maximum number of 6.4 MHz channels that the
CM can support, Y
max
be the maximum number of 3.2 MHz channels that the CM can support, and Z
max
be the
maximum number of 1.6 MHz channels that the CM can support. Z
max
is equivalent to the maximum number of
upstream transmitters that the CM can support.
The CM reports X
max
, Y
max
, and Z
max
during registration using modem capabilities encodings specified in [MULPI],
Annex subsection, "Modem Capabilities Encoding." The CM MUST provide that X
max
Y
max
Z
max
.
Communications of X
max
, Y
max
, and Z
max
by a CM provides enough information to describe all the combinations
possible for X, Y, and Z.
From these values the CMTS may calculate B
max
= max(6.4*X
max
, 3.2*Y
max
, 1.6*Z
max
).
Then, all values for (X, Y, Z) are allowed, which satisfy the four following inequalities:
6.4*X + 3.2*Y + 1.6*Z B
max

X X
max

Y Y
max

X + Y + Z Z
max
.
6.3 Downstream
6.3.1 Downstream Protocol and Interleaving Support
44

The downstream PMD sublayer MUST conform to ITU-T Recommendations J.83, [ITU-T J.83-B], except for
Section B.5.2. Interleaver depths are defined in the Downstream Radio Frequency Specification [DRFI]. The
downstream RF output electrical requirements including downstream frequency plan, interleaver depths, spectrum
format, clock, and symbol requirements are defined in [DRFI]. The CM MUST support the interleaver depths
defined in Table 6-1 of [DRFI]. The CM MAY support the interleaver depths defined in [DRFI].
6.3.2 Downstream Electrical Input to CM
The CM MUST be able to accept any number of signals between one to MDBC simultaneously, located in frequency
intervals that comply with the CMs TB list and Demod List. A CM MUST be capable of being reconfigured to
receive different channels. At a minimum, the CM MUST be able to receive at least four downstream channels,
located independently within an arbitrary 60 MHz window in the downstream frequency band. The CM MUST

44
96 CableLabs
11/17/11
support bonding of any number of downstream channels up to its maximum. The CM MUST be able to locate and
accept RF modulated signals located within channels defined in [CEA-542-B] for Harmonic Related Carrier (HRC),
Incremental Related Carrier (IRC), and Standard (STD) North American frequency plans. Operation below a center
frequency of 111 MHz is not required. The signals will have the characteristics defined in Table 622.
Table 622 - Electrical Input to CM
45

Parameter Value
Center Frequency of DOCSIS
channel
111 to 867 MHz 30 kHz, CM MAY tune 111 to 999 MHZ 30 KHz
Level Range (one DOCSIS channel) -15 dBmV to +15 dBmV
Modulation Type 64-QAM and 256-QAM
Symbol Rate (nominal) 5.056941 Msym/s (64-QAM) and 5.360537 Msym/s (256-QAM)
Bandwidth 6 MHz (alpha = 0.18 Square Root Raised Cosine shaping for 64-QAM and
alpha = 0.12 Square Root Raised Cosine shaping for 256-QAM)
Total Input Power (40 MHz and
above)
< 33 dBmV
Maximum average power of carrier
input to CM, within any 6 MHz
channel from 54 MHz up to 1002
MHz
X = average power of lowest power demodulated QAM channel
Channels demodulated within the CM:
Min (X + 10 dB, +15 dBmV).
Non-demodulated 6 MHz bands within the CM
a) Min (X + 10 dB, +20 dBmV), for zero, one, or two 6 MHz bands.
b) Min (X + 10 dB, +15 dBmV), for all other 6 MHz bands besides the two
possible allocated exceptions
46

Input (load) Impedance 75 ohms
Input Return Loss > 6 dB (across the selected upstream frequency range 5-42 MHz or 5-85
MHz).
> 6 dB (108 to 870 MHz, or 108 to 1002 MHz if CM center frequency ranges
up to 999 MHz)
Connector F connector per [ISO/IEC-61169-24] or [SCTE 02] (common with the output)

6.3.3 CM BER Performance
47

The bit-error-rate performance of a CM is described in this section. For each individual downstream received
channel, the CM MUST meet the BER requirements with the interleaver set for the I = 128, J = 1 mode of
interleaving.
6.3.3.1 64-QAM
6.3.3.1.1 64-QAM CM BER Performance
Implementation loss of the CM MUST be such that the CM achieves a post-FEC BER less than or equal to 10
-8
when
operating at a carrier to noise ratio (E
s
/N
o
) of 23.5 dB or greater. If it is not possible to measure post-FEC BER

45
Table modified per PHYv3.0-N-07.0456-1 #4 on 6/12/07 by KN.
46
Revised this line per PHY3.0-N-07.0389-1 by PO on 5/7/07.
47
Paragraph revised per PHYv3.0-N-07.0458-1 #1 on 6/12/07 by KN.
11/17/11 CableLabs
97
directly, Codeword Error Rate, R
C
(as defined below) may be used. In this case, the CM MUST achieve a Codeword
Error Rate of less than or equal to 9 x 10
-7
when operating at a carrier to noise ratio (E
s
/N
o
) of 23.5 dB or greater.
Computation of Codeword Error Rate, R
C
:
( )
( ) ( ) ( )
0 0 0
0
C C E E E E
E E
R
C C U U
U U
C
+ +

=
Where:
E
U
is the value of the count of code words with uncorrectable errors;
E
C
is the value of the count of code words with correctable errors and;
C is the value of the count of code words without errors.
Sample the values at the beginning of the test interval (indicated by the subscripted values, E
U0
, E
C0
,and C
0
) and the
end of the test interval (indicated by the values, E
U
, E
C
, and C).
6.3.3.1.2 Image Rejection Performance
Performance as described in Section 6.3.3.1.1 MUST be met with analog or digital signal at +10 dBc in any portion
of the RF band other than the adjacent channels.
6.3.3.1.3 64-QAM Adjacent Channel Performance
Performance as described in Section 6.3.3.1.1 MUST be met with a digital signal at 0 dBc in the adjacent channels.
Performance as described in Section 6.3.3.1.1 MUST be met with an analog signal at +10 dBc in the adjacent
channels.
Performance as described in Section 6.3.3.1.1, with an additional 0.2 dB allowance, MUST be met with a digital
signal at +10 dBc in the adjacent channels.
6.3.3.2 256-QAM
6.3.3.2.1 256-QAM CM BER Performance
-8
when
s
/N
o
) as shown below. If it is not possible to measure post-FEC BER directly,
Codeword Error Rate, R
C
(as defined in Section 6.3.3.1.1) may be used. In this case, the CM MUST achieve a
Codeword Error Rate of less than or equal to 9 x 10
-7
s
/N
o
), as shown in
the following:
Input Receive Signal Level E
s
/N
o
-6 dBmV to +15 dBmV 30 dB or greater
Less than -6 dBmV down to -15 dBmV 33 dB or greater
98 CableLabs
11/17/11
6.3.3.2.2 256-QAM Image Rejection Performance
Performance as described in Section 6.3.3.2.1 MUST be met with an analog or a digital signal at +10 dBc in any
portion of the RF band other than the adjacent channels.
6.3.3.2.3 256-QAM Adjacent Channel Performance
Performance as described in Section 6.3.3.2.1 MUST be met with an analog or a digital signal at 0 dBc in the
adjacent channels.
Performance as described in Section 6.3.3.2.1, with an additional 0.5 dB allowance, MUST be met with an analog
Performance as described in Section 6.3.3.2.1, with an additional 1.0 dB allowance, MUST be met with a digital
6.3.4 Downstream Multiple Receiver Capabilities
This section describes the mechanism by which the CM communicates to the CMTS the constraints on receive
channel assignment imposed by the CM's capabilities. The CMTS needs to be aware of the CM's capabilities when it
assigns or changes the parameters of a downstream channel received by a CM. If the CM were reconfigured without
regard to its capabilities, a disruption of downstream data and/or DOCSIS master clock synchronization could result.
The CM reports its capabilities to the CMTS by sending its Receive Channel Profiles (RCP). In response, the CMTS
configures the CM by sending a Receive Channel Configuration (RCC).
The CM MUST support all the capability it reports to the CMTS.
The CMTS MUST NOT command the CM to operate outside its advertised capabilities.
The [MULPI] specification defines the detailed encodings by which Receive Channel Profiles and Receive Channel
Configurations describe Receive Modules, Receive Channels, and the interconnections between them.
6.3.4.1 Receive Module Parameters
A Receive Module (RM) is a grouping of channels, together with parameters describing the constraints on those
channels. Examples of RMs are:
Tuner, with constraints on capture bandwidth and frequency range.
Demodulator, with constraints on contiguous positioning of channels, frequency range, modulation order, and
interleaver depth.
A Receive Channel Profile communicated from CM to CMTS defines the following attributes of each Receive
Module:
Adjacent Channels: The number of contiguous channels processed by the Receive Module.
Channel Block Range: The minimum center frequency of the lowest-frequency channel in the block and the
maximum center frequency of the highest-frequency channel in the block.
Common Physical Layer Parameters: A list of physical layer parameters which are shared by all Receive
Channels connected to the Receive Module. Examples may include modulation type and interleaver settings.
Connection Capability: A list of the higher level (closer to RF port) Receive Modules to which this Receive
Module can connect.
11/17/11 CableLabs
99
A Receive Channel Configuration (RCC), communicated from CMTS to CM, assigns one or more of the following
attributes of a Receive Module:
First Channel Center Frequency: Center frequency of the lowest-frequency channel of a block of adjacent
channels.
Connection Assignment: Specifies to which single higher-level Receive Module to connect.
6.3.4.2 Receive Channel Parameters
A Receive Channel (RC) is the packet stream output to the MAC layer originating from a single QAM channel. A
Receive Channel Profile, communicated from CM to CMTS, defines the following attributes of each Receive
Channel:
Connection Capability: A list of the Receive Modules to which the Receive Channel can connect.
Connected Offset: The offset of the Receive Channel within a block of adjacent channels.
Primary Downstream Channel Capability: A flag indicating whether the Receive Channel is capable of
providing the DOCSIS master clock reference to the CM.
A Receive Channel Configuration communicated from CMTS to CM assigns the following attributes to a Receive
Channel:
Center Frequency Assignment: The RF center frequency of the Receive Channel.
Primary Downstream Channel Indicator: A flag indicating that the CMTS assigns this Receive Channel the
responsibility to provide master clock reference timing to the CM.
Connection Assignment: Specifies to which Receive Module the Receive Channel is to connect.
6.3.4.3 Standard Receive Channel Profile
In order to limit the complexity in the CMTS of configuring arbitrarily complicated Receive Channel Profiles,
DOCSIS defines a set of "Standard" Receive Channel Profiles that describe a minimal set of constraints.
A CM reports to the CMTS at least one standard RCP as well as its "Manufacturer" RCP that gives more details of
its capabilities and constraints. If the CMTS configures the CM with a Receive Channel Configuration based on a
Standard RCP, some CM capabilities represented in the Manufacturer RCP but outside the Standard RCP may not be
available.
The Standard Receive Channel Profile named "6-DOCSIS-01" is defined for 6 MHz operation as depicted in Figure
633. The RF input from the cable plant is connected to a Receive Module with a block of 10 adjacent channels.
This Receive Module represents the constraints of a tuner with a 60 MHz capture bandwidth that can be placed at
any point in the full DOCSIS frequency range. The demodulation of four channels anywhere within the capture
bandwidth is represented by the four Receive Channels, each corresponding to the packet stream from a single QAM
channel output to the MAC layer.
100 CableLabs
11/17/11
Tuner
Receive Module 1
[ Adjacent Channels Constraint: 10 ]
RC1 RC2 RC3 RC4
MAC Sublayer
Receive
Channels
(each cor-
responds to
single QAM
channel)
Packets
RF Port
Diplexer
Splitter
RF Connection to HFC Plant

Figure 633 - 6 MHz DOCSIS Standard Receive Channel Profile 01
For this Standard Receive Channel Profile, the CMTS sends the following parameters to the CM in the Receive
Channel Configuration:
48

Center frequency of the first channel in the 10-channel Receive Module.
Center frequency of each Receive Channel within the capture bandwidth of the Receive Module.
6.3.4.4 Example Manufacturer Receive Channel Profile
A more complex example showing a Manufacturer Receive Channel Profile as reported by the CM to the CMTS is
depicted in Figure 634. The RF input from the cable plant is distributed to three Receive Modules, corresponding to
tuner modules with capture bandwidth constraints. A network connects the tuners to a group of four Receive
Modules, corresponding to demodulators with contiguous channel constraints. Each demodulator connects to four
Receive Channels, each representing a single QAM channel output to the MAC layer.
In response, the CMTS configures each Receive Module and Receive Channel as follows:
Frequency range of each analog tuner.
Center frequency of first channel of each demodulator.
Center frequency of each Receive Channel.

48
Channels constraint changed from 8 to 10 in Figure 6-33 per PHY3.0-N-07.0408-1, PO 5-7-07
11/17/11 CableLabs
101
Interconnection between Receive Modules.
Interconnection between lower-level (demodulator) Receive Modules and Receive Channels.
RF Port Diplexer Splitter
Tuner
Receive Module 1
Tuner
Receive Module 2
Tuner
Receive Module 3
Demodulator
Receive Module 4
Demodulator
Receive Module 5
Demodulator
Receive Module 6
Demodulator
Receive Module 7
Interconnection Network
RC1
RC2
RC3
RC4
RC5
RC6
RC7
RC8
RC9
RC10
RC11
RC12
RC13
RC14
RC15
RC16
MAC Sublayer
Receive Channels
(each corresponds
to single QAM
channel)
Packets

Figure 634 - Example Manufacturer Receive Channel Profile

6.3.4.5 Reconfiguration of CM Receive Channels
Before the CMTS sends a DBC-REQ (dynamic bonding change request) asking the CM to change parameters of a
receive channel, the CMTS can determine the constraints and connectivity of the CM based on the Receive Channel
Profile and Receive Channel Configuration. From this information, the CMTS can determine whether the requested
change can be implemented by the CM, and whether the change will interrupt other downstream channels. In this
way, the CMTS can minimize and/or schedule the disruption of downstream traffic and any perturbations to DOCSIS
master clock timing resulting from receive channel parameter changes.
6.3.5 Non-Synchronous DS Channel Support
The "primary downstream channel" of a CM is defined as the downstream channel from which it derives CMTS
master clock timing for upstream transmission. All other concurrently received channels are called "secondary
downstream channels."
102 CableLabs
11/17/11
A DOCSIS 3.0 CM MUST support secondary downstream channels that are asynchronous to the primary
downstream channel, i.e., with a different long-term average QAM symbol rate, but that conform to all DOCSIS 3.0
specifications.
11/17/11 CableLabs
103
Annex A Timing Requirements for Supporting Business Services
over DOCSIS
The CMTS transmits its 10.24 MHz reference clock timing to the CM via the downstream QAM symbol clock and
via timing information embedded in downstream SYNC messages. The CM utilizes this downstream clock
information to regenerate a local 10.24 MHz reference clock. The CM reference clock, together with ranging
commands, is used to accurately time upstream TDMA or S-CDMA burst transmissions.
49

A.1 CMTS
50

DOCSIS 3.0 CMTSs have the following timing and synchronization requirements:
The CMTS MUST be able to lock the downstream DOCSIS symbol clock to an external Stratum 1 clock
source.
The CMTS MUST be able to lock the downstream DOCSIS SYNC messages to an external Stratum 1 clock
source, as defined by the CMTS Timestamp Jitter requirements in [DRFI].
Examples of Stratum 1 clock sources are the Primary Reference Source (PRS) clock and the clock source obtained
from a DTI timing server [DTI].
A.2 CM
The requirement for a DOCSIS 3.0 CMTS to support a symbol clock that is traceable to a Stratum 1 reference
introduces the possibility for a CM to pass that precise clock on to subtending devices in the customer premises.
Support for such a clock output is outside the scope of this specification. One example of the usage of such a clock
output is given in [TEI], where the CM outputs the recovered master clock to a "Clocking Unit" in an embedded
TDM Emulation Adapter which uses the traceable clock to provide T1 or E1 emulation service.
51

49
50
Revised this section per PHYv3.0-N-06.0342-1 by GO on 1/12/07.
51
Replaced this section per PHYv3.0-N-06.0342-1 by GO on 1/12/07.
104 CableLabs
11/17/11
Annex B Additions and Modifications for European Specification
This annex applies to the second technology option referred to in Section 1.1. For the first option, refer to Sections 5
and 6.
This annex defines the interface for the physical layer used in conjunction with the European multi-program
television distribution. It describes the electrical characteristics and signal processing operations for what is generally
called EuroDOCSIS Cable Modem (CM) and Cable Modem Termination System (CMTS). This is an optional annex
and in no way affects certification of equipment adhering to the North American technology option described in the
sections referenced above.
Requirements referring to the absolute level of electrical power are represented in units of dBmV throughout this
specification. This is in contrast to other European standards applicable in cabled distribution systems for television
and sound signals (e.g., [EN 50083-7] or [EN 50083-10]) where such requirements are expressed traditionally in
units of dBV. The values in dBmV used in this specification can be transformed into values in dBV by adding 60
dB.
The numbering of the paragraphs in this annex has been maintained such that the suffix after the letter for the annex
refers to the part of the specification where the described changes apply to. In cases where the requirements for both
technology options are identical, a reference is provided to the main text.
B.1 Scope
B.1.1 Introduction and Purpose
See Section 1.1.
B.1.2 Background
B.1.2.1 Broadband Access Network
A coaxial-based broadband access network is assumed. This may take the form of either an all-coax or hybrid
fiber/coax (HFC) network. The generic term "cable network" is used here to cover all cases.
A cable network uses a shared-medium, tree-and-branch architecture, with analog transmission. The key functional
characteristics assumed in this document are the following:
Two-way transmission.
A maximum optical/electrical spacing between the CMTS and the most distant CM of 160 km (route meters) in
each direction.
A maximum differential optical/electrical spacing between the CMTS and the closest and most distant modems
of 160 km (route meters) in each direction.
At a propagation velocity in fiber of approximately 5 ns/m, 160 km of fiber in each direction results in a round-trip
delay of approximately 1.6 ms.
B.1.2.2 Network and System Architecture
B.1.2.2.1 The DOCSIS Network
See Section 1.2.2.1.
11/17/11 CableLabs
105
B.1.2.3 Service Goals
See Section 1.2.3.
B.1.2.4 Statement of Compatibility
This annex specifies an interface, commonly referred to as EuroDOCSIS 3.0, which is the third generation of the
interface, (earlier versions are commonly referred to as EuroDOCSIS 1.x and 2.0). EuroDOCSIS 3.0 MUST be
backward- and forward-compatible with equipment built to the previous specifications. EuroDOCSIS 3.0-compliant
CMs MUST interoperate seamlessly with EuroDOCSIS 2.0 and EuroDOCSIS 1.x CMTSs, albeit in the 2.0 and 1.x
modes, as the case may be. EuroDOCSIS 3.0-compliant CMTSs MUST seamlessly support EuroDOCSIS 2.0 and
EuroDOCSIS 1.x CMs.
B.1.2.5 Reference Architecture
See Section 1.2.5.
B.1.2.6 DOCSIS 3.0 Documents
A list of the specifications in the DOCSIS 3.0 series is provided in Table B1. If additions and modifications are
required to define the European technology option, they are included in a normative annex to each of the
specifications. Otherwise, the listed specifications equally apply to EuroDOCSIS 3.0. For further information, please
refer to http://www.cablemodem.com.
Table B1 - DOCSIS 3.0 Series of Specifications
Designation Title
CM-SP-PHYv3.0 Physical Layer Specification
CM-SP-MULPIv3.0 Media Access Control and Upper Layer Protocols Interface Specification
CM-SP-OSSIv3.0 Operations Support System Interface Specification
CM-SP-SECv3.0 Security Specification

This annex is defining the interface for the physical layer of the European technology option.
B.1.3 Requirements
See Section 1.3.
B.1.4 Conventions
See Section 1.4.
B.1.5 Organization of Document
See Section 1.5.
B.2 References
B.2.1 Normative References
See Section 2.1.
106 CableLabs
11/17/11
B.2.2 Informative References
See Section 2.2.
B.2.3 Reference Acquisition
See Section 2.3.
B.3 Terms and Definitions
See Section 3.
B.4 Abbreviations and Acronyms
See Section 4.
B.5 Functional Assumptions
This section describes the characteristics of a cable television plant, assumed to be for the purpose of operating a
data-over-cable system. It is not a description of CMTS or CM parameters. The data-over-cable system MUST be
interoperable within the environment described in this section.
Whenever a reference in this section to frequency plans or to compatibility with other services conflicts with any
legal requirement for the area of operation, the latter shall take precedence. Any reference to analog TV signals in a
particular frequency band does not imply that such signals are physically present.
B.5.1 Equipment Assumptions
B.5.1.1 Frequency Plan
In the downstream direction, the cable system is assumed to have a pass band with a lower edge of down to 47 MHz.
Typically, the lower edge of the downstream pass band is at 87.5 MHz. The upper edge is implementation-dependent
but is typically in the range of 300 to 862 MHz. Within that pass band, PAL/SECAM analog television signals in 7/8
MHz channels and FM radio signals are assumed to be present, as well as other narrowband and wideband digital
signals. 8 MHz channels are used for data communication.
In the upstream direction, the cable system is assumed to have a pass band of 5-65 MHz. PAL/SECAM analog
television signals in 7/8 MHz channels may be present, as well as other signals.
B.5.1.2 Compatibility with Other Services
See Section 5.1.2.
B.5.1.3 Fault Isolation Impact on Other Users
See Section 5.1.3.
B.5.1.4 Cable System Terminal Devices
Compliance with EMC requirements is not covered by this specification. The protection requirements with respect to
electromagnetic compatibility are contained in harmonized standards published in the Official Journal of the
European Union.
11/17/11 CableLabs
107
Any reference in the present document to the transmission of television in the forward channel that is not consistent
with [EN 300 429] is outside the normative scope as only [EN 300 429] is used for digital multi-program TV
distribution by cable in European applications.
Requirements for safety are outside the scope of the present document. Safety standards for European applications
are published by CENELEC. Examples of such CENELEC product safety standards are [EN 60950-1] and [EN
50083-1]. For CENELEC safety categories of interfaces, see [EG 201 212].
B.5.2 RF Channel Assumptions
See Section 5.2.
B.5.2.1 Transmission Downstream
The RF channel transmission characteristics of the cable network in the downstream direction assumed for the
purposes of minimal operating capability are described in Table B2. These numbers assume total average power of
a digital signal in an 8 MHz channel bandwidth for carrier levels unless indicated otherwise. For impairment levels,
the numbers in Table B2 assume average power in a bandwidth in which the impairment levels are measured in a
standard manner for cable TV systems. For analog signal levels, the numbers in Table B2 assume nominal analog
video carrier level (peak envelope power) in a 7/8 MHz channel bandwidth. Transmission is from head-end
combiner to the CM input at the customer location. All conditions are present concurrently.
Table B2 - Assumed Downstream RF Channel Transmission Characteristics
Parameter Value
Frequency range Cable system downstream operating range is from 47 MHz
to 862 MHz. However, the operating range for data
communication is from 108 to 862 MHz and the values in
this table only apply to this frequency range. The use of
frequencies between 108 and 136 MHz may be forbidden
due to national regulation with regard to interference with
aeronautical navigation frequencies.
RF channel spacing (design bandwidth) 7/8 MHz, 8 MHz channels are used for data communication
Carrier-to-noise ratio in an 8 MHz band (analog video
level)
Not less than 44 dB
1

Carrier-to-interference ratio for total power (discrete
and broadband ingress signals)
Not less than 52 dB within the design bandwidth
Composite triple beat distortion for analog modulated
carriers
Not greater than -57 dBc within the design bandwidth
2

Composite second-order distortion for analog
modulated carriers
Not greater than -57 dBc within the design bandwidth
2

Cross-modulation level Under consideration
Amplitude ripple 2.5 dB in 8 MHz
Group delay ripple in the spectrum occupied by the
CMTS
100 ns over frequency range 0.5 4.43 MHz
Micro-reflections bound for dominant echo -10 dBc @ 0.5 s
-15 dBc @ 1.0 s
-20 dBc @ 1.5 s
-31.5 dBc @ > 1.5 s
108 CableLabs
11/17/11
Parameter Value
Burst noise Not longer than 25 s at a 10 Hz average rate
Seasonal and diurnal signal level variation 8 dB
Signal level slope, 85 862 MHz Maximum slope of 12 dB in either the positive or negative
direction
Maximum analog video carrier level at the system
outlet, inclusive of above signal level variation
17 dBmV
3

Lowest analog video carrier level at the system outlet,
inclusive of above signal level variation
0 dBmV
4

Table Notes:
Note 1 This presumes that the average digital carrier is operated at analog peak carrier level. When the digital
carrier is operated below the analog peak carrier level, this C/N may be less
Note 2 For SECAM systems the value is no greater than -52 dBc within the design bandwidth
Note 3 For SECAM systems the value is 14 dBmV
Note 4 For SECAM systems the value is -3 dBmV

B.5.2.2 Transmission Upstream
The RF channel transmission characteristics of the cable network in the upstream direction assumed for the purposes
of minimal operating capability are described in Table B3. Transmission is from the CM output at the customer
location to the head-end. All conditions are present concurrently.
52

Table B3 - Assumed Upstream RF Channel Transmission Characteristics
Parameter Value
Frequency range. 5 to 65 MHz edge to edge
Carrier-to-noise ratio in active channel Not less than 22 dB
Carrier-to-ingress power (the sum of discrete and
broadband ingress signals) ratio in active channel
Carrier-to-interference (the sum of noise, distortion,
common-path distortion and cross-modulation) ratio in
active channel
Burst noise Not longer than 10 s at a 1 KHz average rate for most
cases
2,3

Amplitude ripple across upstream operating frequency
range (maximum)
2.5 dB in 2 MHz
Group delay ripple across upstream operating
frequency range (maximum)
300 ns in 2 MHz

52
11/17/11 CableLabs
109
Parameter Value
Micro-reflections (maximum) single echo -10 dBc @ 0.5 s
-20 dBc @ 1.0 s
-31.5 dBc @ > 1.0 s
Seasonal and diurnal signal level variation Not greater than 12 dB min to max
Table Notes:
Note 1 Ingress avoidance or tolerance techniques may be used to ensure operation in the presence of time-
varying discrete ingress signals that could be as high as 0 dBc.
Note 2 Amplitude and frequency characteristics sufficiently strong to partially or wholly mask the data carrier.
Note 3 Impulse noise levels more prevalent at lower frequencies (<15 MHz).

B.5.2.2.1 Availability
B.5.3 Transmission Levels
See Section 5.3.
B.5.4 Frequency Inversion
See Section 5.4.
B.6 Physical Media Dependent Sublayer Specification
B.6.1 Scope
This section defines the electrical characteristics and signal processing operations for a cable modem (CM) and
Cable Modem Termination System (CMTS). It is the intent of this specification to define an interoperable CM and
CMTS such that any implementation of a CM can work with any CMTS. It is not the intent of this specification to
imply any specific implementation.
This section applies to the second technology option referred to in Section 1. In cases where the requirements for
both technology options are identical, a reference is provided to the main text.
B.6.2 Upstream
B.6.2.1 Overview
The upstream Physical Media Dependent (PMD) sublayer uses an FDMA/TDMA (herein called TDMA mode) or
FDMA/TDMA/S-CDMA (herein called S-CDMA mode) burst type format, which provides six modulation rates and
multiple modulation formats. The use of TDMA or S-CDMA mode is configured by the CMTS via MAC messaging.
FDMA (frequency division multiple access) indicates that multiple RF channels are assigned in the upstream band. A
CM transmits on one or more RF channels and may be reconfigured to change channels.
A CM MUST support at least 4 active upstream channels (which are referred to as the Transmit Channel Set for that
CM).
110 CableLabs
11/17/11
The CM reports its maximum number of upstream channels capability, and certain other of its capability
characteristics, to the CMTS (Section B.6.2.25).
The CM MUST be able to operate each channel in the Transmit Channel Set, simultaneously, anywhere in the
upstream band, subject to restrictions on transmit power across the channels and on reconfiguration of certain
transmit properties (see Sections B.6.2.19 and B.6.2.20, and subsections of each). The CMTS MUST be capable of
assigning and receiving each RF channel anywhere in the upstream band. The CMTS MUST set the number of
assigned channels and assigned channels' center frequency, and all other channel attributes. The CMTS MAY change
the number of assigned channels and the channel attributes. Each RF channel has its own set of UCD parameters as
defined in [MULPI].
TDMA (time division multiple access) indicates that upstream transmissions have a burst nature. A given RF channel
is shared by multiple CMs via the dynamic assignment of time slots. S-CDMA (synchronous code division multiple
access) indicates that multiple CMs can transmit simultaneously on the same RF channel and during the same TDMA
time slot, while being separated by different orthogonal codes.
In this document, the following naming conventions are used. For TDMA, the term "modulation rate" refers to the
RF channel symbol rate (160 to 5120 ksym/s). For S-CDMA, the term "modulation rate" refers to the "chip rate,"
which is the rate (1280 to 5120 kHz) of the individual elements (chips) of the S-CDMA spreading code. Modulation
rates are represented in units of "Hz" denoting the number of symbols per second in TDMA mode or the number of
chips per second in S-CDMA mode. The "modulation interval" is the symbol period (TDMA mode) or chip period
(S-CDMA mode) and is the reciprocal of the modulation rate. At the output of the spreader, a group of 128 chips
which comprise a single S-CDMA spreading code, and are the result of spreading a single information (QAM
constellation) symbol is referred to as a "spread symbol." The period of a spread symbol (128 chips) is called a
"spreading interval." A "burst" is a physical RF transmission that contains a single preamble plus data, and (in the
absence of preceding and following bursts) exhibits RF energy ramp-up and ramp-down.
In some cases logical zeros or logical ones are used to pad data blocks; this indicates data with zero-valued or one-
valued binary bits, which result in non-zero transmitted RF energy. In other cases a numerical zero is used; this
denotes, for example, symbols which result in zero transmitted RF energy (after ramp-up and ramp-down are taken
into account).
The modulation format includes pulse shaping for spectral efficiency, is carrier-frequency agile, and has selectable
output power level.
Each burst supports a flexible modulation order, modulation rate, preamble, randomization of the payload, and
programmable FEC encoding.
All of the upstream transmission parameters associated with burst transmission outputs from the CM are configurable
by the CMTS via MAC messaging. Many of the parameters are programmable on a burst-by-burst basis.
The PMD sublayer can support a near-continuous mode of transmission, wherein ramp-down of one burst MAY
overlap the ramp-up of the following burst, so that the transmitted envelope is never zero. In TDMA mode, the
system timing of the TDMA transmissions from the various CMs MUST provide that the center of the last symbol of
one burst and the center of the first symbol of the preamble of an immediately following burst are separated by at
least the duration of five symbols. The guard band MUST be greater than or equal to the duration of five symbols
plus the maximum timing error. Timing error is contributed by both the CM and CMTS. CM timing performance is
specified in Section B.6.2.20.1. Maximum timing error and guard band may vary with CMTSs from different
vendors. The term guard time is similar to the guard band, except that it is measured from the end of the last symbol
of one burst to the beginning of the first symbol of the preamble of an immediately following burst. Thus, the guard
time is equal to the guard band 1.
The PMD sublayer also supports a synchronous mode of transmission when using S-CDMA, wherein ramp-down of
one burst MAY completely overlap the ramp-up of the following burst, so that the transmitted envelope is never
11/17/11 CableLabs
111
zero. There is no guard time for transmission on S-CDMA channels. The system timing of the S-CDMA
transmissions from the various CMs MUST provide adequate timing accuracy so that different CMs do not
appreciably interfere with each other. S-CDMA utilizes precise synchronization so that multiple CMs can transmit
simultaneously.
The upstream modulator is part of the cable modem which interfaces with the cable network. The modulator contains
the electrical-level modulation function and the digital signal-processing function; the latter provides the FEC,
preamble prepend, symbol mapping, and other processing steps.
At the Demodulator, similar to the Modulator, there are two basic functional components: the demodulation function
and the signal processing function. The Demodulator resides in the CMTS and there is one demodulation function
(not necessarily an actual physical demodulator) for each carrier frequency in use. The demodulation function
receives all bursts on a given frequency.
The demodulation function of the Demodulator accepts a varying-level signal centered around a commanded power
level and performs symbol timing and carrier recovery and tracking, burst acquisition, and demodulation.
Additionally, the demodulation function provides an estimate of burst timing relative to a reference edge, an estimate
of received signal power, may provide an estimate of signal-to-noise ratio, and may engage adaptive equalization to
mitigate the effects of a) echoes in the cable plant, b) narrowband ingress and c) group delay. The signal-processing
function of the Demodulator performs the inverse processing of the signal-processing function of the Modulator.
This includes accepting the demodulated burst data stream and decoding, etc. The signal-processing function also
provides the edge-timing reference and gating-enable signal to the demodulators to activate the burst acquisition for
each assigned burst slot. The signal-processing function may also provide an indication of successful decoding,
decoding error, or fail-to-decode for each codeword and the number of corrected Reed-Solomon symbols in each
codeword. For every upstream burst, the CMTS has a prior knowledge of the exact burst length in modulation
intervals. (See Sections B.6.2.5, B.6.2.5.1, B.6.2.5.2, B.6.2.6, B.6.2.20, and in the MAC Service ID section of
[MULPI].)
B.6.2.2 Signal Processing Requirements
See Section 6.2.2.
B.6.2.3 Modulation Formats
See Section 6.2.3.
B.6.2.4 R-S Encode
B.6.2.4.1 R-S Encode Modes
B.6.2.4.2 R-S Bit-to-Symbol Ordering
B.6.2.5 Upstream R-S Frame Structure for DOCSIS 3.0 Multiple Transmit Channel mode
Enabled
See Section 6.2.5.
B.6.2.5.1 R-S Codeword Length
112 CableLabs
11/17/11
B.6.2.5.1.1 Burst Size
See Section 6.2.5.1.1.
B.6.2.5.1.2 Fixed Codeword Length
B.6.2.5.1.3 Shortened Last Codeword
B.6.2.5.2 R-S FEC Disabled
B.6.2.6 Upstream R-S Frame Structure for DOCSIS 3.0 Multiple Transmit Channel mode Not
Enabled
See Section 6.2.6.
B.6.2.7 TDMA Byte Interleaver
See Section 6.2.7.
B.6.2.7.1 Byte Interleaver Parameters
B.6.2.7.2 Interleaver Operating Modes
B.6.2.7.2.1 Fixed Mode
B.6.2.7.2.2 Dynamic Mode
B.6.2.8 Scrambler (Randomizer)
See Section 6.2.8.
B.6.2.9 TCM Encoder
See Section 6.2.9.
B.6.2.9.1 Byte to TCM Symbol Mapping
11/17/11 CableLabs
113
B.6.2.10 Preamble Prepend
See Section 6.2.10.
B.6.2.11 Modulation Rates
See Section 6.2.11.
B.6.2.11.1 DOCSIS 3.0 Modulation Rates
B.6.2.11.2 Backward Compatibility Modulation Rates
B.6.2.12 S-CDMA Framer and Interleaver
B.6.2.12.1 S-CDMA Framing Considerations
B.6.2.12.2 Mini-slot Numbering
B.6.2.12.2.1 Mini-slot Numbering Parameters in UCD
B.6.2.12.2.2 Mini-slot Numbering Examples
B.6.2.12.3 Transmission Time
B.6.2.12.4 Latency Considerations
B.6.2.12.5 Spreader-off Bursts for Maintenance on S-CDMA channel
B.6.2.12.6 Limiting the Number of Codes Assigned to a CM
B.6.2.13 S-CDMA Framer
See Section 6.2.13.
114 CableLabs
11/17/11
B.6.2.13.1 Subframe Definition
B.6.2.13.2 Framer Operation
B.6.2.13.2.1 Rules for Preamble and Coded TCM Symbols
See Section B.6.2.13.2.1.
B.6.2.13.2.2 Rules for Uncoded Symbols and the Uncoded TCM Subsymbols
B.6.2.13.2.3 Subframe Example
B.6.2.13.2.4 Frame Transmission
B.6.2.14 Symbol Mapping
See Section 6.2.14.
B.6.2.15 S-CDMA Spreader
See Section 6.2.15.
B.6.2.15.1 Code Hopping
B.6.2.15.1.1 Code Hopping Mode 1
B.6.2.15.1.2 Code Hopping Mode 2
B.6.2.15.1.3 Code Hopping Generator
B.6.2.16 Transmit Pre-Equalizer
A transmit pre-equalizer of a linear equalizer structure, as shown in Figure B1, MUST be configured for each
upstream channel used by the CM in response to the Ranging Response (RNG-RSP) message transmitted by the
CMTS.
11/17/11 CableLabs
115
There are two modes of operation for the pre-equalizer of a CM: DOCSIS 1.1 mode, and DOCSIS 2.0 pre-
equalization mode: In DOCSIS 1.1 mode, the CM MUST support a (T)-spaced equalizer structure, with 8 taps. The
pre-equalizer MAY have 1, 2, or 4 samples per symbol, with a tap length longer than 8 symbols: In DOCSIS 1.1 pre-
equalization mode, for backwards compatibility, the CMTS MAY support fractionally spaced equalizer format (T/2
and T/4). In DOCSIS 2.0 pre-equalization mode, the pre-equalizer MUST support a symbol (T)-spaced equalizer
structure with 24 taps.
In DOCSIS 1.x-only logical channels (type 1 logical channels [MULPI]), the CM and the CMTS MUST use
DOCSIS 1.1 pre-equalization mode.
In DOCSIS 2.0-only or DOCSIS 3.0-only logical channels (type 3 or type 4 logical channels [MULPI]), the CM and
the CMTS MUST use DOCSIS 2.0 pre-equalization mode.
In DOCSIS 1.x/2.0 mixed logical channels (type 2 logical channels [MULPI]), the CM and the CMTS MUST use
DOCSIS 1.1 pre-equalization mode from initial ranging until DOCSIS 2.0 mode is enabled or until Multiple
Transmit Channel mode is enabled in the registration process (if either is enabled). The CM and CMTS MUST use
DOCSIS 2.0 pre-equalization mode after DOCSIS 2.0 is enabled for the CM. In the case that a CM is placed into
Multiple Transmit Channel mode, the CM and CMTS use DOCSIS 2.0 pre-equalization for all bursts on type 2
logical channels whether the burst descriptor in the UCD for the transmitted burst is a Type 4 or Type 5 burst
descriptor.
+
2 3
I-QInput
z
-1
z
-1
z
-1
z
-1
z
-1
z
-1
+ + + + +
... ...
... ...
F
2
F
1
F
4
F
3
F
17
F
18
F
24
Equalizer
Output

Figure B1 - Transmit Pre-Equalizer Structure

The RNG-RSP MAC message carries the CM transmit equalization information, and may instruct the CM to either
convolve the equalizer coefficients, or (in DOCSIS 2.0 pre-equalization mode only) load them directly. When the
CM is instructed to convolve the transmit equalizer coefficients, it MUST convolve the coefficients sent by the
CMTS in the RNG-RSP with the existing coefficients, to get the new coefficients. After convolving, the CM MUST
truncate the convolution result, such that 24 taps (8 taps in DOCSIS 1.1 pre-equalization mode) remain after the
truncation, with the main tap located at the tap designated by the last RNG-RSP received by the CM. The operation
of the convolution is formulized by the following equation:
( )
( )
24 ... 1 ,
1
1 1
1 1
1
^
1 , 24 min
24 . 1 max
1
= =
+
+ +
+ +
+ +
+
+ =
+
+
n F F F
m
m m m
m m m
m m l k
L L n L
L L n L k
m
L L k n
m
n
where:
m
n
F are the coefficients prior to the convolution;
116 CableLabs
11/17/11
1 + m
n
F are the coefficients after the convolution;
n
F
^
are the coefficients sent from the CMTS;
m
L
is the main tap location prior to the convolution;
1 + m
L is the main tap location after the convolution as dictated by the CMTS.
In DOCSIS 2.0 pre-equalization mode, the CMTS MAY instruct the CM to load the transmit pre-equalizer
coefficients. When the CM is instructed to load the transmit equalizer coefficients, it MUST load the coefficients
sent by the CMTS into the pre-equalizer coefficients after proper normalization, if necessary.
In DOCSIS 1.x-only logical channels, in response to an initial ranging request and periodic ranging requests prior to
CM registration, when the CMTS sends the pre-equalizer coefficients, the CMTS MUST compute and send them
with an equalizer length of 8 and in T-spaced format, where T is the modulation interval. After registration, the
CMTS MAY use a fractionally spaced equalizer format (T/2- or T/4-spaced), with a longer tap length, to match the
CM pre-equalizer capabilities that the CMTS learned from the REG-REQ message modem capabilities field.
In DOCSIS 2.0-only or DOCSIS 3.0-only logical channels, the CMTS MUST compute and send the pre-equalizer
coefficients with an equalizer length of 24, and in T-spaced format, at all times.
In DOCSIS 1.x/2.0 mixed logical channels, in response to an initial ranging request and periodic ranging requests
prior to CM registration, when the CMTS sends the pre-equalizer coefficients, the CMTS MUST compute and send
them with an equalizer length of 8 and in T-spaced format. After registration, if the DOCSIS 1.1 pre-equalization
mode is enabled, the CMTS MAY use a fractionally spaced equalizer format (T/2- or T/4-spaced), with a longer tap
length, to match the CM pre-equalizer capabilities that the CMTS learned from the REG-REQ message modem
capabilities field. If DOCSIS 2.0 pre-equalization mode or Multiple Transmit Channel mode is enabled for the CM,
the CMTS MUST use a T-spaced equalizer structure with 24 taps. If the first update of the pre-equalizer after the
activation of DOCSIS 2.0 pre-equalization mode uses "convolve" mode, the CM MUST zero-pad the existing 8-tap
filter to a 24-tap filter, and convolve, according to the rules above.
Prior to making an initial ranging request, and whenever the upstream channel frequency or upstream channel
modulation rate changes, the CM MUST initialize the coefficients of the pre-equalizer to a default setting in which
all coefficients are zero, except the real coefficient of the first tap (i.e., F1). Whenever the main location is changed,
the CM, not the CMTS, MUST compensate for the delay (ranging offset), due to a shift from the previous main tap
location, to a new main tap location of the equalizer coefficients, sent by the CMTS (in both "convolve" and "load"
operations). The pre-equalizer coefficients are then updated through the subsequent ranging process (unicast initial
ranging and periodic ranging).
In DOCSIS 1.1 pre-equalization mode, the CMTS MUST NOT move the main tap location during periodic ranging.
In DOCSIS 1.1 pre-equalization mode, the CMTS MUST NOT instruct the CM to load the transmit equalizer
coefficients.
In DOCSIS 2.0 pre-equalization mode, the CMTS MAY move the main tap location during unicast initial ranging or
periodic ranging.
Equalizer coefficients may be included in every RNG-RSP message, but typically they only occur when the CMTS
determines that the channel response has significantly changed. The frequency of equalizer coefficient updates in the
RNG-RSP message is determined by the CMTS.
11/17/11 CableLabs
117
The CM MUST normalize the transmit equalizer coefficients in order to guarantee proper operation (such as not to
overflow or clip). The CM MUST NOT change its target transmit power due to gain or loss of the new coefficients
in both "convolve" and "load" operations. The target power is defined in Section B.6.2.19.
In DOCSIS 1.1 mode, if the CM equalizer structure implements the same number of coefficients as assigned in the
RNG-RSP message, then the CM MUST NOT change the location of the main tap in the RNG-RSP message. If the
CM equalizer structure implements a different number of coefficients than defined in the RNG-RSP message, the
CM MAY shift the location of the main tap value. The CM MUST adjust its ranging offset, in addition to any
adjustment in the RNG-RSP message, by an amount that compensates for the movement of the main tap location.
B.6.2.17 Spectral Shaping
See Section 6.2.17.
B.6.2.17.1 Upstream Frequency Agility and Range
The CM MUST support operation over the Standard Upstream Frequency Range of 5-65 MHz, edge to edge.
Additionally, the CM MAY support selectable operation over a) the Standard Upstream Frequency Range of 5-65
MHz, or b) the Extended Upstream Frequency Range of 5-85 MHz, edge to edge.
53

Offset frequency resolution MUST be supported per Table B9.
B.6.2.17.2 Spectrum Format
B.6.2.18 Relative Processing Delays
See Section 6.2.18.
B.6.2.19 Transmit Power Requirements
54

The following requirements apply with Multiple Transmit Channel mode enabled. Requirements with Multiple
Transmit Channel mode disabled are addressed in Section B.6.2.19.4 below.
The CM MUST support varying the amount of transmit power. Requirements are presented for 1) range of reported
transmit power per channel; 2) step size of power commands; 3) step size accuracy (actual change in output power
per channel compared to commanded change); and 4) absolute accuracy of CM output power per channel. The
protocol by which power adjustments are performed is defined in [MULPI]. Such adjustments by the CM MUST be
within the ranges of tolerances described below. A CM MUST confirm that the transmit power per channel limits are
met after a RNG-RSP is received for each of the CM's active channels that are referenced. An active channel for a
CM is defined as any channel for which the CM has been given a grant, which it will use for channel initialization, or
ranging, or any channel for which the CM is "ranged." The set of "active channels" is also called the Transmit
Channel Set. (There will be some mechanism by which a CMTS can command a CM to deactivate an active channel,
for the purpose of either reducing the number of active channels on that CM or for the purpose of adding a different
active channel as a replacement upstream channel for that CM.) Note that the set of channels actually bursting
upstream from a CM is a subset of the active channels on that CM; often one or all active channels on a CM will not
be bursting, but such quiet channels are still "active channels" for that CM.
Transmit power per channel is defined as the average RF power in the occupied bandwidth (channel width)
transmitted in the data symbols of a burst, assuming equal likely QAM symbols, measured at the F-connector of the
CM. Total transmit power is defined as the sum of the transmit power per channel of each channel transmitting a

53
Inserted per PHYv3.0-N-11.1021-1 on 11/4/11 by PO.
54
Section revised per PHYv3.0-N-07.0570-2 on 1/25/08 and PHYv3.0-N-08.0661-3 on 4/10/08 by KN.
118 CableLabs
11/17/11
burst at a given time. Maximum and minimum transmit power per channel level requirements refer to the CM's target
transmit power per channel levels, defined as the CM's estimate of its actual transmit power per channel. The actual
transmitted power per channel MUST be within 2 dB of the target power. The target transmit power per channel
MUST be variable over the range specified in Table B9. The target transmit power per channel MAY be variable
over a range that extends above the maximum levels specified in Table B4, Table B5, and Table B6. The CM
MUST NOT exceed a target transmit power level per channel of 61 dBmV. Note that all fidelity requirements
specified in Section B.6.2.22 still apply when the CM is operating over an extended transmit power range.
The CM communicates its ability to transmit above the maximum levels defined in Table B4, Table B5, and Table
B6 to the CMTS via a modem capability encoding as defined in [MULPI]. When the CM indicates that it supports
the extended range and the CMTS confirms this capability, for each combination of the number of channels in the
Transmit Channel Set and constellation the CM and CMTS MUST use the greater of the value from the CMs
advertised capability encoding and the default value of Pmax specified in Table B4, Table B5, and Table B6 as
Pmax for that particular combination. When the CM indicates that it supports the extended range and the CMTS
disables this capability, the CM and CMTS MUST use the default values of Pmax specified in Table B4, Table B
5, and Table B6.
55

With Multiple Transmit Channel mode enabled, let P
load
= P
hi
P
r
, for each channel, using the definitions for P
hi
and
P
r
in the following subsections of B.6.2.19. The channel corresponding to the minimum value of P
load
is called the
highest loaded channel, and its value is denoted as P
load_1
in this specification, even if there is only one channel in the
Transmit Channel Set. A channel with high loading has a low P
load_n
value (but not less than 0); the value of P
load_n
is
analogous to an amount of back-off for an amplifier from its max power output. A channel has lower power output
when that channel has a lower loading (more back-off) and thus a higher value of P
load_n
. Note that the highest loaded
channel is not necessarily the channel with the highest transmit power, since a channels max power depends on the
modulations it is supporting in its burst profiles. The channel with the second lowest value of P
load
is denoted as the
second highest loaded channel, and its loading value is denoted as P
load_2
. The channel with the third lowest value of
P
load
is the third highest loaded channel, and its loading value is denoted as P
load_3
; and the channel with the fourth
lowest loading value is denoted as the fourth highest loaded channel, and its loading value is P
load_4
(in the case of
fewer than 4 channels in the Transmit Channel Set, P
load_n
will only be valid for the n active upstream channels).
P
load_min_set
defines the upper end of the Dynamic Range Window for the CM with respect to P
hi
for each channel.
P
load_min_set
will limit the maximum power possible for each active channel to a value less than P
hi
when P
load_min_set
is
greater than zero. P
load_min_set
is a value commanded to the CM from the CMTS when the CM is given a TCC in
registration and RNG-RSP messages [MULPI]. P
load_min_set
, P
load_n,
P
hi_n
, P
r
_
n,
etc. are defined only when Multiple
Transmit Channel mode is enabled.
56

The CMTS MUST provide P
load_min_set
such that P
hi_n
P
load_min_set
P
low_n
for each active channel, or
equivalently,
0 P
load_min_set
P
hi_n
P
low_n

CMTS commands adjusting P
r_n
MUST be consistent with the P
load_min_set
previously assigned to the CM and with the
following limits:
P
load_min_set
P
hi_n
P
r_n
P
load_min_set
+ 12 dB
and the equivalent:
P
hi_n
(P
load_min_set
+ 12 dB) P
r_n
P
hi_n
P
load_min_set
57

55
56
57
Added this equation per PHYv3.0-N-06.0340-1 by GO on 1/11/07.
11/17/11 CableLabs
119
A value is computed, P
low_multi_n
, for each channel in the Transmit Channel Set, which sets the lower end of the
transmit power Dynamic Range Window for that channel, given the upper end of the range (which is determined by
P
load_min_set
).
58

P
low_multi_n
= P
hi_n
P
load_min_set
12 dB.
The effect of P
low_multi n
is to restrict the dynamic range required (or even allowed) by a CM across its multiple
channels, when operating with multiple active channels.
load_min_set
load_min_set
to P
r_n
th
changes to P
r_n
load_min_set
. The CMTS SHOULD NOT issue
a change in P
load_min_set
r_n
until after also providing a sufficient reconfiguration time
on the n
th
channel. The CMTS SHOULD NOT issue a change in P
load_min_set
after commanding a prior change in
P
load_min_set
until after also providing a global reconfiguration time for the first command. Also, the CMTS SHOULD
NOT issue a change in P
r_n
a) until after providing a global reconfiguration time following a command for a new
value of P
load_min_set
, and b) until after providing a sufficient reconfiguration time on the n
th
channel after issuing a
previous change in P
r_n
r_n
and/or P
load_min_set

to the CM without a sufficient reconfiguration time for instituting the first command. When a concurrent new value
of P
load_min_set
and change in P
r_n
are commanded, the CM MAY wait to apply the change in P
r_n
at the next global
load_min_set
change at the first sufficient reconfiguration time of the n
th
channel.
59

Note that the CMTS can allow small changes of power in the CM's highest loaded channel, without these fluctuations
impacting the transmit power dynamic range with each such small change. This is accomplished by setting P
load_min_set

to a smaller value than normal, and fluctuation of the power per channel in the highest loaded channel is expected to
wander. The CM also MUST ignore any commanded change of per channel transmit power which would result in
P
r_n
falling below the Dynamic Range Window, i.e., lower power than permitted by the Dynamic Range Window,
which is determined by P
load_min_set
. The CM also MUST ignore a command changing P
load_min_set
such that existing
values of P
r_n
would fall outside the new Dynamic Range Window.
The following two paragraphs define the CM and CMTS behavior in cases where there are Dynamic Range Window
violations due to indirect changes to P
hi_n
or addition of a new channel with incompatible parameters, without direct
change of P
r_n
or P
load_min_set
.
A change in the number of active channels changes P
hi_n
of the existing active channels (due to different P
hi_n

calculation tables for different number of active channels). Prior to changing the number of active channels, the
CMTS SHOULD change P
r_n
of all current active channels, if necessary, to fit in the new expected Dynamic Range
Window. If the change in the number of active channels would cause a violation of the Dynamic Range Window of
one of the existing active channels, the CM will ignore the change (i.e., not adding/removing an active channel). In
such case, the CM will also indicate partial service message in the REG-ACK or DBC response [MULPI].
When adding a new active channel to the transmit channel set, the new channel's power is calculated according to the
offset value define in TLV 46.8.4 [MULPI], if it is provided. The CMTS SHOULD NOT set an offset value that will
result in a P
r_n
for the new channel outside the Dynamic Range Window. In the absence of the TLV, the new
channels power is initially set by the CM at the minimum allowable power, i.e., the bottom of the Dynamic Range
Window.

58
Revised this sentence and the following equation per PHYv3.0-N-06.0340-1.
59
Revised this paragraph and added new requirement per PHYv3.0-N-06.0341-1 by GO on 1/12/07.
120 CableLabs
11/17/11
If one of the new active channels added by the change in the number of active channels violates the Dynamic Range
Window, the CM will ignore the change (i.e., not adding/removing an active channel). In such case, the CM will also
indicate partial service message in the REG-ACK or DBC response [MULPI].
For a UCD change, P
hi_n
may change due to a change in the modulation types. Additionally, a UCD change that
modifies the symbol rate may result in a change in P
low_n
. The CMTS MUST NOT send a UCD change for the n
th

active channel that violates P
hi_n
P
load_min_set
P
low_n
. If a UCD changes the P
hi_n
for the n
th
active channel [MULPI],
the CM sets P
r_n
= P
hi_n
- P
load_n
for that channel, when the change count of the MAP matches the change count in the
new UCD. The new (changed) P
hi_n
and the current P
load_n
are used in the calculation. The CM maintains P
load_n
at its
value at the time the MAP change count matches the new UCD change count, and calculates a new P
r_n
and target
power for the channel, to be applied for bursts granted in the MAP with change count matching the change count in
the new UCD. The change in the transmit power P
r_n
due to the recalculation with the new P
hi_n
is not considered a
change in the commanded transmit power, for example as in Section B.6.2.19, Section B.6.2.20, and Section
B.6.2.22, and thus this change is required to be transparent on all active channels continuously bursting throughout
such a change. The CM is expected to meet all requirements without need of global reconfiguration time, even with
the induced change in P
r_n
, unless, of course, other parameters are changed which do imply the CM should be
provided a global reconfiguration time.
60

The spurious performance requirements of Sections B.6.2.22.1, B.6.2.22.1.1, and B.6.2.22.1.2 apply when the CM is
operating within certain ranges of values for P
load_n
, for n = 1 to the active number of upstream channels, and for
certain ranges of the number of spreading codes being transmitted as a fraction of the total number of active codes on
a channel for S-CDMA channels, as detailed in those sections.
Transmit power per channel, for each channel, as reported by the CM in the MIB, is referenced to the 64-QAM
constellation. When transmitting with other constellations, a slightly different transmit power will result, depending
on the constellation gain in Table B4, Table B5, and Table B6 below. As an example, if the reported power in a
channel is 30 dBmV, 64-QAM will be transmitted with a target power of 30 dBmV in that channel, while QPSK will
be transmitted with 28.82 dBmV in that channel.
Table B4 - Constellation Gains and Per Channel Power Limits with One Channel in the Transmit Channel
Set
Constellation
Constellation
Gain Gconst
Relative to
64-QAM (dB)
Pmin
(dBmV) Pmax
(dBmV)
TDMA
Pmax
(dBmV)
S-CDMA
Pmin - Gconst
(dBmV)
Pmax -
Gconst
(dBmV)
TDMA
Pmax - Gconst
(dBmV)
S-CDMA
L M H
QPSK -1.18 17 20 23 61 56 18.18 62.18 57.18
8-QAM -0.21 17 20 23 58 56 17.21 58.21 56.21
16-QAM -0.21 17 20 23 58 56 17.21 58.21 56.21
32-QAM 0.00 17 20 23 57 56 17.00 57.00 56.00
64-QAM 0.00 17 20 23 57 56 17.00 57.00 56.00
128-QAM 0.05 17 20 23 N/A 56 16.95 N/A 55.95

60
Paragraph replaced per PHYv3.0-N-08.0661-3 on 4/10/08 by KN.
11/17/11 CableLabs
121
Table B5 - Constellation Gains and Per Channel Power Limits with Two Channels in the Transmit Channel
Set
Constellation
Constellation
Gain Gconst
Relative to
64-QAM (dB)
Pmin
(dBmV) Pmax
(dBmV)
TDMA
Pmax
(dBmV)
S-CDMA
Pmin - Gconst
(dBmV)
Pmax -
Gconst
(dBmV)
TDMA
Pmax -
Gconst
(dBmV)
S-CDMA
L M H
QPSK -1.18 17 20 23 58 53 18.18 59.18 54.18
8-QAM -0.21 17 20 23 55 53 17.21 55.21 53.21
16-QAM -0.21 17 20 23 55 53 17.21 55.21 53.21
32-QAM 0.00 17 20 23 54 53 17.00 54.00 53.00
64-QAM 0.00 17 20 23 54 53 17.00 54.00 53.00
128-QAM 0.05 17 20 23 N/A 53 16.95 N/A 52.95

Table B6 - Constellation Gains and Per Channel Power Limits with Three or Four Channels in the Transmit
Channel Set
Constellation
Constellation
Gain Gconst
Relative to
64-QAM (dB)
Pmin
(dBmV)
Pmax
(dBmV)
TDMA
Pmax
(dBmV)
S-CDMA
Pmin - Gconst
(dBmV)
Pmax -
Gconst
(dBmV)
TDMA
Pmax -
Gconst
(dBmV)
S-CDMA
L M H
QPSK -1.18 17 20 23 55 53 18.18 56.18 54.18
8-QAM -0.21 17 20 23 52 53 17.21 52.21 53.21
16-QAM -0.21 17 20 23 52 53 17.21 52.21 53.21
32-QAM 0.00 17 20 23 51 53 17.00 51.00 53.00
64-QAM 0.00 17 20 23 51 53 17.00 51.00 53.00
128-QAM 0.05 17 20 23 N/A 53 16.95 N/A 52.95

The actual transmitted power per channel, within a burst, MUST be constant to within 0.1 dB peak to peak, even in
the presence of power changes on other active channels. This excludes the amplitude variation, theoretically present,
due to QAM amplitude modulation, pulse shaping, pre-equalization, and for S-CDMA, spreading and varying
number of allocated codes.
The CM MUST support the transmit power calculations defined in Section B.6.2.19.1 and Section B.6.2.19.2.
B.6.2.19.1 TDMA Transmit Power Calculations
61

In TDMA mode, the CM determines its target transmit power per channel P
t
as follows, for each channel that is
active. Define for each active channel, for example, upstream channel X:
P
r
= Reported power level (dBmV) of CM in MIB (refers to 64-QAM constellation) for channel X.
P = Power level adjustment (dB); for example, as commanded in ranging response message.
G
const
= Constellation gain (dB) relative to 64-QAM constellation per Section B.6.2.19.
Pmin = Minimum target transmit power per channel permitted for the CM per Section B.6.2.19.

61
Section revised per PHYv3.0-N-07.0459-2 on 6/12/07 and PHYv3.0-N-08.0611-3 on 4/10/08 by KN.
122 CableLabs
11/17/11
P
max
= Maximum target transmit power permitted for the CM per Section B.6.2.19.
Phi = min(P
max
- G
const
) over all burst profiles used by the CM in channel X per Section B.6.2.19.
Plow = max(P
min
- G
const
) over all burst profiles used by the CM in channel X per Section B.6.2.19.
P
t
= Target transmit power per channel level (dBmV) of CM in channel X (actual transmitted power per channel
in channel X as estimated by CM).
hi
r
= P
hi
- P
load
for that
hi
for the
load
1. P
r
= P
r
2. P
r
= min[P
r
, P
hi
3. P
r
= max[P
r
, P
low
4. Test, IF[P
r
 P
hi
-P
load_min_set
r
The CM then transmits, in channel X with target power per channel P
t
= P
r
+ G
const
, i.e., the reported power, plus the
constellation gain.
Usually the reported power level is a relatively constant quantity, while the transmitted power level in channel X
varies dynamically as different burst profiles, with different constellation gains, are transmitted. A CM's target
transmit power per channel MUST never be below P
min
or above P
max
. This implies that in some cases, the extreme
transmit power levels (e.g., 61 dBmV for QPSK and 17 dBmV) may not be permitted if burst profiles with multiple
constellations are active. Also, if only QPSK is used, the reported power per channel may be greater than 61 dBmV,
although the target transmit power per channel will not exceed 61 dBmV.
For example, if with two channels active in channel X, only QPSK and 64-QAM burst profiles are active, P
hi
= 54
dBmV and P
low
low
permitted QPSK transmitted power in channel X is 54 dBmV - 1.2 dB = 52.8 dBmV, the minimum QPSK power in
channel X is 18.2 dBmV - 1.2 dB = 17 dBmV (with 1280 kHz modulation rate), the maximum 64-QAM power in
channel X is 54 dBmV, and the minimum 64-QAM power in channel X is 18.2 dBmV (with 1280 kHz modulation
rate).
B.6.2.19.2 S-CDMA Transmit Power Calculations
11/17/11 CableLabs
123
B.6.2.19.2.1 S-CDMA Transmit power Calculations with Maximum Scheduled Codes Not Enabled
62

In S-CDMA mode when Maximum Scheduled Codes is not enabled, the CM determines its target transmit power per
channel P
t
as follows, for each channel which is active. Define for each active channel, (for example, upstream
channel X):
P
r
= reported power level (dBmV) of CM in MIB (refers to 64-QAM constellation and all active codes
transmitted) for channel X.
P
hi
= min[P
max
- G
const
] over all burst profiles used by the CM in channel X per Section B.6.2.19.
P
low
= max[P
min
- G
const
is over all burst profiles used by the CM in channel X per Section B.6.2.19.
hi
r
= P
hi
- P
load
for that
hi
for the
load
1. P
r
= P
r
2. P
r
= min[P
r
, P
hi
3. P
r
= max[P
r
, P
low
4. Test, IF[P
r
 P
hi
-P
load_min_set
r
In a spreader-on frame, the CM then transmits each code i with target power:
P
t,i
= P
r
+ G
const,i
- 10 log(number_active_codes)
i.e., the reported power for channel X, plus the constellation gain G
const,i
of that code, less a factor taking into account
the number of active codes. The total transmit power in channel X, P
t
, in a frame is the sum of the individual transmit
powers P
t,i
of each code in channel X, where the sum is performed using absolute power quantities [non-dB domain].
In a spreader-off frame, the CM target transmit power in channel X is P
t
= P
r
+ G
const
.
The transmitted power level in channel X varies dynamically as the number of allocated codes varies, and as
different burst profiles with different constellation gains, are transmitted. A CM's target transmit power per channel
MUST never be below P
min
or above P
max
, including over all numbers of allocated codes and all burst profiles. This
implies that in some cases the extreme transmit power levels per channel (e.g., 17 and 56 dBmV) may not be
permitted. Also, if only QPSK is used, the reported power in a channel may be greater than 56 dBmV, although the
target transmit power per channel will not exceed 56 dBmV.
If, for example, the Transmit Channel Set contains only one channel, and this channel, channel X, has QPSK and 64-
QAM burst profiles, the number of active codes is 128 and the number of codes per mini-slot is 2, then P
hi
= 56
dBmV and P
low
low
permitted QPSK transmitted power in channel X is 56 1.18 = 54.82 dBmV when all active codes are transmitted.
The minimum QPSK power in channel X is 36.24 dBmV 1.18 dB 10log(128) dB + 10log(2) dB = 17 dBmV

62
Section revised per PHYv3.0-N-07.0459-2 on 6/12/07 and per PHYv3.0-N-08.0611-3 on 4/10/08 by KN.
124 CableLabs
11/17/11
(with 1280 kHz modulation rate), when one mini-slot is transmitted. The last term in the sum is the result of
summing the individual powers over 2 codes. Similarly, the maximum 64-QAM power in channel X is 56 dBmV,
when all active codes are transmitted and the minimum 64-QAM power in channel X is 36.24 dBmV 10log(128)
dB + 10log(2) dB = 18.18 dBmV (with 1280 kHz modulation rate) when one mini-slot is transmitted. The minimum
QPSK power permitted in channel X while transmitting, for example, 2 mini-slots is 20 dBmV (with 1280 kHz
modulation rate), and the minimum 64-QAM power permitted in channel X, while transmitting 2 mini-slots, is 21.2
dBmV (with 1280 kHz modulation rate).
The CM needs to implement some form of clipping on the transmitted waveform at the higher output powers in order
to prevent peak to average ratio (PAR) issues.
The power received at the CMTS in a spreader-on frame will sometimes be less than the nominal power of a
spreader-off frame because of such factors as: 1) broadcast opportunities not used by any CM; 2) unicast grants not
used by one or more CMs; or 3) mini-slots assigned to the NULL SID.
B.6.2.19.2.2 S-CDMA Transmit Power Calculations With Maximum Scheduled Codes Enabled.
63

In S-CDMA mode on channels on which Maximum Scheduled Codes is enabled, the CM determines its target
transmit power per channel P
t
as follows, for each channel which is active. Define for each upstream channel, for
example, upstream channel X:
P
r
= reported power level (dBmV) of CM in MIB (operational transmit power of the spreader-off ranging burst
referenced to 64-QAM modulation) for channel X.
P
hi_S
= min[P
max
- G
const
] over all spreader-on burst profiles used by the CM in channel X per Section B.6.2.19.
P
low_S =
max[17 - G
const
is over all burst profiles used by the CM in channel X per Section B.6.2.19.
P
max_T
= Maximum target transmit power permitted for the CM in channel X in TDMA mode per Section
B.6.2.19 for the constellation used in ranging.
P
hi_T
= min[P
max
_
T
- G
const
] over all spreader-off burst profiles used by the CM in channel X per Section
B.6.2.19.
P
on
= P
r
clipped at the maximum spreader-on limit.
P
sf
= CM Power Shortfall.
P
hr
= S-CDMA Power Headroom in dB. Equivalent to the value of the Ranging Response Message TLV-11
divided by 4.
P = power level adjustment in dB sent from CMTS to CM for channel X.
hi
r
= P
hi
- P
load
for that
hi
for the
load
64

The CM updates its power per channel in each channel by the following steps:
1. P
r
= P
r

63
Section modified per PHYv3.0-N-07.0459-2 on 6/12/07 and PHYv3.0-N-08.0611-3 on 4/10/08 by KN.
64
11/17/11 CableLabs
125
2. P
r
= min[P
r
, P
hi_T
] //Clip at max TDMA power limit per channel.
3. P
r
= max[P
r
, P
low_S
] //Clip at min S-CDMA power limit per channel.
4. Test, IF[P
r
 P
hi_T
-P
load_min_set
r
6. P
on
= min[P
r
, P
hi_S
-P
load_min_set
] //Clip at max S-CDMA power limit per channel for the multiple channel dynamic
range limitation, as set by P
load_min_set
.
In spreader-off frames, the CM transmits in channel off with target power:
P
t
= P
r
+ G
const

Based on the spreader-off transmit power in channel X, the CM updates its power shortfall in channel X according to
the following steps:
with only one channel in the Transmit Channel Set:
P
sf
= P
t
- (56 - P
load_min_set
) // Difference between spreader-off and max spreader-on target powers in channel
X;

P
sf
= max[P
sf
, 0] // Set P
sf
to 0 if P
t
load_min_set
in channel X;
and with more than one channel in the Transmit Channel Set:
P
sf
= P
t
- (53 - P
load_min_set
) // Difference between spreader-off and max spreader-on target powers in channel
X;
P
sf
= max[P
sf
, 0] // Set P
sf
to 0 if P
t
load_min_set
in channel X.
In spreader-on frames, the CM transmits each code i with target power:
P
t, i
= P
on
+ G
const
,
i
- 10 log(number_active_codes) + P
hr

i.e., the clipped reported power in channel X plus the constellation gain G
const
,
i
of that code, less a factor taking into
account the number of active codes, plus the Power Headroom P
hr
. P
hr
is the power (in dB) added to account for CMs
that have maximum scheduled code limits and can transmit additional power per code. The total transmit power in
channel X, P
t,i
in a frame is the sum of the individual transmit powers P
t
,
i
of each code in channel X, where the sum
is performed over all N
alloc
allocated codes using absolute power quantities (non-dB domain).

If, for example, with more than one channel in the Transmit Channel Set, and the burst profile in channel X contains
QPSK for IUCs 1, 2, 3, and 4 and 64-QAM for IUCs 9 and 10, the number of active codes is 128, and the number of
codes per mini-slot is 2, then P
hi_S
= 53 dBmV, P
low
_
S
= 36.24 dBmV, and P
hi
_
T
= 58 dBmV. Assume P
load_min_set
= 0
dB. Assume the CM ranges channel X at spreader-off target transmit power of 57 dBmV. The CM reports for
channel X P
sf
= 57 dBmV - 53 dBmV = 4 dB. The CMTS uses P
sf
to set for channel X (using its vendor-specific
algorithm) max_scheduled_codes = 32 and P
hr
= 6 dB. (The S-CDMA power headroom may differ from the power
shortfall, at the discretion of the CMTS.) The CM sets its transmitted power per code in channel X to:
126 CableLabs
11/17/11
P
t, i
= P
on
+ G
const
,
i
- 10 log (number_active_codes) + P
hr

= 53 dBmV + 0 dB - 21 dB + 6 dB // For a code with 64-QAM modulation in channel X
= 38 dBmV
A parameter that may be used to illustrate the effect of increased power per code in channel X is the Effective
Transmit Power, P
eff
, the power that would result in channel X hypothetically if all N
a
active codes were transmitted.
It is computed as:

where the last term is the average constellation gain.
For a reference case with all codes transmitted in channel X using 64-QAM modulation (G
const
= 0 dB), the effective
transmit power reduces to:
P
eff
= P
on
+ P
hr

Continuing the above example, the result is:
P
eff
= 53 dBmV + 6 dB
= 59 dBmV
Limiting the number of codes has given the CM an enhanced effective power of 59 dBmV, which is 6 dB above the
normal maximum power per channel of 53 dBmV, and 2 dB above the ranging power of 57 dBmV. In this example,
the CMTS used its discretion to ask for 2 dB more enhancement in channel X than was needed (P
hr
= 6 dB vs P
sf
= 4
dB), perhaps due to some known impairment in the channel.
The effective SNR is an SNR estimate for a given code corresponding to the effective transmit power in channel X.
It is defined as the measured SNR at the last station maintenance, minus the CM power shortfall, plus the power
headroom, plus the difference in constellation gain between the ranging burst and the code under consideration. Its
equation is:
sf
+ P
hr
+ (G
const
,
i
- G
const
,
ranging
).
where G
const
,
ranging
is the constellation gain of the ranging burst in channel X that resulted in the SNR measurement.
In the MIB, the effective SNR corresponds to a reference case with 64-QAM modulation (G
const, i
= 0 dB):
sf
+ P
hr
- G
const
,
ranging.

Continuing the example, if the measured SNR in channel X in the last station maintenance was 17 dB, using QPSK
modulation (G
const
,
ranging
= -1.2 dB), then the effective SNR referenced to 64-QAM modulation is:
effective_SNR = 17dB - 4dB + 6dB + 1.2 dB = 20.2 dB.
B.6.2.19.3 Transmit Power Step Size
11/17/11 CableLabs
127
B.6.2.19.4 Transmit Power Requirements for Multiple Transmit Channel mode Not Enabled
With Multiple Transmit Channel mode disabled, the CM MUST operate as in [DOCSIS RFI 2.0] with the following
exceptions.
The minimum upstream power P
min
MUST be as given in Table B7 below, superseding the minimum upstream
power requirements in [DOCSIS RFI 2.0].
The CM MUST replace the value of 53 in the equation for P
hi
_S in [DOCSIS RFI 2.0] with 56 since this value
is actually intended to be Pmax from Table B4 above.
Maximum upstream power prior to registration:
If the Extended Upstream Transmit Power Support TLV in the MDD is OFF, the CM MUST use a
maximum upstream power Pmax as given in Table B4 above, superseding the maximum upstream power
requirements in [DOCSIS RFI 2.0].
If the Extended Upstream Transmit Power Support TLV in the MDD is ON, the CM MAY exceed the
maximum upstream power Pmax as given in Table B4 above. The CM MUST NOT exceed a target
transmit power level of 61 dBmV. Note that all fidelity requirements specified in Section B.6.2.22 still
apply when the CM is operating over an extended transmit power range.
Maximum upstream power after registration:
The CM MUST use a maximum upstream power Pmax as given in Table B4 above, superseding the
maximum upstream power requirements in [DOCSIS RFI 2.0].
65

Table B7 - Minimum Transmit Power Pmin for Multiple Transmit Channel mode Not Enabled
Modulation Rate (kHz) P
min
(dBmV) Applicability
1 280 17 CM MUST support

B.6.2.20 Burst Profiles
The transmission characteristics are separated into three portions: a) Channel Parameters; b) Burst Profile Attributes;
and c) User Unique Parameters. The Channel Parameters include: a) the modulation rate (six rates from 160 kHz to
5120 kHz in octave steps); b) the center frequency (Hz); c) the 1536-bit Preamble Superstring; and d) the S-CDMA
channel parameters. The Channel Parameters are shared by all users on a given channel. The Burst Profile Attributes
are listed in Table B8, and these parameters are the shared attributes corresponding to a burst type.
The CM MUST generate each burst at the appropriate time, as conveyed in the mini-slot grants provided by the
CMTS MAPs.
The CM MUST support all burst profiles commanded by the CMTS, via the Burst Descriptors in the UCD, and
subsequently assigned for transmission in a MAP.

65
Revised paragraph per PHYv3.0-N-06.0341-1 by GO on 1/12/07 and per PHYv3.0-N-10.0942-5 on 9/22/10 by JB.
128 CableLabs
11/17/11
Table B8 - Burst Profile Attributes
66

Modulation QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, 128-QAM(TCM Only)
Differential Encoding On/Off
TCM Encoding On/Off
Preamble Length 0-1536 bits (note Section B.6.2.10)
Preamble Value offset 0 to 1534
R-S FEC Error Correction (T) 0 to 16 (0 implies no R-S FEC. The number of codeword parity bytes is
2*T)
R-S FEC Codeword Information Bytes (k) Fixed: 16 to 253 (assuming R-S FEC on)
Shortened: 16 to 253 (assuming R-S FEC on)
Scrambler Seed 15 bits
Maximum Burst Size (mini-slots)
1
0 to 255
Guard Time Size 4 to 255 modulation intervals
There is no guard time in S-CDMA
Last Codeword Length Fixed, shortened
Scrambler On/Off On/Off
R-S Interleaver Depth (I
r
)
2
0 to floor(2048/N
r
)
3

R-S Interleaver Block Size (B
r
)
4
2*N
r
to 2048
Preamble Type QPSK0/QPSK1
S-CDMA Spreader
5
On/Off
S-CDMA Codes per Subframe
5
1 to 128
S-CDMA Framer Interleaver Step Size
5
1 to (spreading intervals per frame - 1)
Table Notes:
Note 1 A max burst length of 0 mini-slots in the Burst Profile means that the grant size is not constrained by the
burst profile parameter [MULPI].
Note 2 If depth=1, no interleaving; if depth=0, dynamic mode.
Note 3 Nr is the R-S codeword size k+2T as defined in Section B.6.2.5.1.
Note 4 Used only in dynamic mode
Note 5 Used only for S-CDMA channels.

The User Unique Parameters may vary for each user even when using the same burst type on the same channel as
another user (for example, Power Level), and are listed in Table B9.

66
Table revised per PHYv3.0-N-07.0459-2 #s 14 & 15 on 6/12/07 and PHYv3.0-N-07.0598-1 on 1/25/08 by KN.
11/17/11 CableLabs
129
Table B9 - User Unique Burst Parameters
67

Power Level
1
8-bit two's complement,
resolution = 0.25 dB
TDMA:
+17 to +57 dBmV (32-QAM, 64-QAM)
+17 to +58 dBmV (8-QAM, 16-QAM)
+17 to + 61 dBmV (QPSK)
S-CDMA: +17 to +56 dBmV
(all modulations)
Resolution = 0.25 dB or better
Offset Frequency
2
Range = 32 kHz, resolution = 1
Hz
Frequency Range per Section B.6.2.17.1
Upstream Frequency Agility and Range
Ranging Offset Integer part: 32-bit two's
complement, resolution = (1 /
10.24 MHz) = 6.25 s/64 =
97.65625 ns

Fractional part: unsigned 8-bit
fractional extension, resolution =
6.25 s/(64*256) =
0.3814697265625 ns
Range: sufficient for maximum cable
plant length per Section B.1.2.1
Broadband Access Network

Resolution:Non-synchronous:
6.25s/64.
Synchronous: 6.25 s/(64*256)
68

Burst Length (mini-slots) if variable
on this channel (changes burst-to-
burst)
N/A 1 to 255 mini-slots
Transmit Equalizer Coefficients
(See Section B.6.2.16, Transmit Pre-
Equalizer)
DOCSIS 3.0 & 2.0 modes: 24
coefficient
(2 real and 2 imaginary), load and
convolve modes

DOCSIS 1.1 mode: up to 64
coefficient (2 real and 2
imaginary), convolve mode only
DOCSIS 3.0 & 2.0 modes: 24 complex
coefficients

DOCSIS 1.1 mode: up to 64 complex
coefficients
Table Notes:
Note 1 The minimum power level limit depends on modulation rate. The maximum power level limit depends on
modulation order and the number of upstream channels.
Note 2 The CM MUST implement the Offset Frequency Adjustment to effect a change in upstream carrier
frequency within 10 Hz of the commanded change.

B.6.2.20.1 Ranging Offset

67
Table modified per PHYv3.0-N-07.0459-2 # 16 on 6/12/07 by KN.
68
Column in row modified per PHYv3.0-N-06.0301-2, PO 11/20/06.
130 CableLabs
11/17/11
B.6.2.20.2 TDMA Reconfiguration Times
69

The CM MUST be capable of switching burst profiles with no reconfiguration time required between bursts, except
for changes in the following parameters: 1) Output Power; 2) Symbol Rate; 3) Offset Frequency; 4) Channel
Frequency; 5) Ranging Offset; and 6) Pre-Equalizer Coefficients. In the case of Multiple Transmit Channel mode
In the case of Multiple Transmit Channel mode enabled, global reconfiguration time (simultaneous on all of the
active channels) could be required by the CM when the Dynamic Range Window is adjusted (new value of
P
load_min_set
), or when any active channel accumulates a transmit power change of more than 3 dB (excluding transmit
hi
[MULPI]) since the last global reconfiguration time (see Section
B.6.2.22.1). The CM MUST meet the requirements, on each active channel, for "Dynamic Range Window change"
(below) whenever P
load_min_set
is changed or re-commanded.
For Dynamic Range Window change or re-command: If P
load_min_set
is changed, or re-commanded, the CM MUST
be able to implement the change between bursts, as long as the CMTS allocates at least 96 symbols, plus 10 s,
between the last symbol center of one burst (on any active channel) and the first symbol center of the following burst
(on any active channel -- same channel or different channel from the previous burst). (Global "quiet" across all active
channels requires the intersection of ungranted burst intervals across all active TDMA channels to be at least 96
symbols plus 10 s duration.) The Output Power of the CM MUST be settled to within 0.1 dB of its final output
power level within 10 s from the beginning of the change in P
load_min_set
, and from the beginning of any Output
Power change on any channel that accompanies the change or re-command of P
load_min_set
. The Dynamic Range
Window, i.e., P
load_min_set
, MUST NOT be changed until the CM is provided sufficient time between bursts by the
CMTS. The Dynamic Range Window MUST NOT change while more than -30 dB of any symbol's energy of the
previous burst remains to be transmitted, or more than -30 dB of any symbol's energy of the next burst has been
transmitted.
For Output Power changes (excluding transmit power changes due to UCD-induced change in P
hi
when Multiple
Transmit Channel mode is enabled [MULPI]): If Output Power is to be changed by 1 dB or less, the CM MUST be
able to implement the change between bursts as long as the CMTS allocates at least 96 symbols, plus 5 s between
the last symbol center of one burst and the first symbol center of the following burst. If Output Power is to be
changed by more than 1 dB, the CM MUST be able to implement the change between bursts, as long as the CMTS
allocates at least 96 symbols, plus 10 s, between the last symbol center of one burst and the first symbol center of
the following burst. The maximum reconfiguration time of 96 symbols should compensate for the ramp-down time of
delay and pre-equalizer delay. The Output Power of the CM MUST be settled to within 0.1 dB of its final output
power level; a) within 5 s from the beginning of a change of 1 dB or less; and b) within 10 s from the beginning of
a change of greater than 1 dB. Output Power MUST NOT be changed until the CM is provided sufficient time
between bursts by the CMTS. Output Power MUST NOT change while more than -30 dB of any symbol's energy of
the previous burst remains to be transmitted, or more than -30 dB of any symbol's energy of the next burst has been
transmitted.
For Symbol Rate changes: The CM MUST be able to transmit consecutive bursts as long as the CMTS allows the
required time between bursts for UCD parameter changes. Symbol Rate MUST NOT be changed until the CM is
provided sufficient time between bursts by the CMTS. Symbol Rate MUST NOT change while more than -30 dB of
the next burst has been transmitted.
For Offset Frequency changes: The CM MUST be able to transmit consecutive bursts as long as the CMTS
allocates at least 96 symbols in between the last symbol center of one burst and the first symbol center of the
following burst. The maximum reconfiguration time of 96 symbols should compensate for the ramp-down time of

69
11/17/11 CableLabs
131
delay and optional pre-equalizer delay. Offset frequency MUST NOT be changed until the CM is provided sufficient
time between bursts by the CMTS. Offset frequency MUST NOT change while more than -30 dB of any symbol's
energy of the previous burst remains to be transmitted, or more than -30 dB of any symbol's energy of the next burst
has been transmitted.
For Channel Frequency changes: The CM MUST be able to implement the change between bursts as long as the
CMTS allocates at least 96 symbols plus 100 ms between the last symbol center of one burst and the first symbol of
the following burst. The Channel Frequency of the CM MUST be settled within the phase noise and accuracy
requirements of Section B.6.2.22.5 and Section B.6.2.22.6 within 100 ms from the beginning of the change. Channel
Frequency MUST NOT be changed until the CM is provided sufficient time between bursts by the CMTS. Channel
Frequency MUST NOT change while more than -30 dB of any symbol's energy of the previous burst remains to be
transmitted, or more than -30 dB of any symbol's energy of the next burst has been transmitted.
For Ranging Offset and/or Pre-Equalizer Coefficient changes: The CM MUST be able to transmit consecutive
bursts as long as the CMTS allocates at least 96 symbols in between the last symbol center of one burst and the first
symbol center of the following burst. The maximum reconfiguration time of 96 symbols should compensate for the
ramp-down time of one burst and the ramp-up time of the next burst, as well as the overall transmitter delay time,
including the pipeline delay and pre-equalizer delay. Ranging Offset MUST NOT be changed until the CM is
provided sufficient time between bursts by the CMTS. Ranging Offset MUST NOT change while more than -30 dB
of any symbol's energy of the previous burst remains to be transmitted, or more than -30 dB of any symbol's energy
of the next burst has been transmitted.
For Modulation Type changes: The CM MUST be able to transmit consecutive bursts with no reconfiguration time
between them (except for the minimum guard time). The modulation MUST NOT change while more than -30 dB of
the next burst has been transmitted, EXCLUDING the effect of the transmit equalizer (if present in the CM). (This is
to be verified with the transmit equalizer providing no filtering; delay only. Note that if the CMTS has decision
feedback in its equalizer, it may need to provide more than the 96 symbol gap between bursts of different modulation
type that the same CM may use; this is a CMTS decision.)
B.6.2.20.3 S-CDMA Reconfiguration Times
70

In S-CDMA mode, for changes in Output Power per mini-slot, Offset Frequency, Pre-equalizer coefficients, and/or
Ranging Offset, the CM MUST be able to transmit consecutive bursts as long as the CMTS allocates the time
duration of at least one frame in between the bursts. For all other burst profile parameter changes, no reconfiguration
is required beyond what is provided by the MAC for such changes. In the case of Multiple Transmit Channel mode
In the case of Multiple Transmit Channel mode enabled, global reconfiguration time could be required by the CM
when the Dynamic Range Window is adjusted (new value of P
load_min_set
), or when any active channel accumulates a
transmit power change of more than 3 dB (excluding transmit power changes due to UCD-induced change in P
hi

[MULPI]) since the last global reconfiguration time (see Section B.6.2.22.1). (Global "quiet" across all active
channels requires the intersection of ungranted burst intervals across all active S-CDMA channels to be at least one
frame.) The CM MUST be able to transmit consecutive bursts, even with a change or re-command of P
load_min_set
, as
long as the CMTS allocates at least one frame in between bursts, across all channels in the Transmit Channel Set.
(From the end of a burst on one channel to the beginning of the next burst on any channel, there must be at least one
frame duration, to provide a "global reconfiguration time" for S-CDMA channels.)
B.6.2.20.4 CM Timing Offsets When Changing Modulation Rate

70
132 CableLabs
11/17/11
B.6.2.21 Burst Timing Convention
See Section 6.2.21.
B.6.2.22 Fidelity Requirements
See Section 6.2.22.
B.6.2.22.1 Spurious Emissions
The noise and spurious power MUST NOT exceed the levels given in Table B10, Table B11, and Table B12.
The noise and spurious power SHOULD NOT exceed the levels given in Table B13 and Table B14. When
Multiple Transmit Channel mode is enabled, the noise and spurious power requirements for the individual channels
are summed (absolute power, NOT in dB) to determine the composite noise floor for the multi-channel transmission
condition. Also when Multiple Transmit Channel mode is enabled, these spurious performance requirements only
apply when the CM is operating within certain ranges of values for P
load_n
, for n = 1 to the number of upstream
channels in the Transmit Channel Set, and for the ratio of number_allocated_codes / number_active_codes from 1
down to 1/4th, for S-CDMA channels.
For S-CDMA mode, when a modem is transmitting fewer than 1/4th of the active spreading codes on a channel, the
spurious emissions requirement limit is the power value (in dBmV), corresponding to the specifications for the
power level associated with 1/4th of the active spreading codes (i.e., 6 dB below the channel power corresponding to
all active codes for that channel allocated to the CM).
When Multiple Transmit Channel mode is enabled and there are two or more channels in the Transmit Channel Set
(TCS), the spurious performance requirements MUST be met only when the following loading ranges are satisfied:
Two or more channels in the TCS: applicable when P
load_1
P
load_min_set
3 dB, AND
Two channels in the TCS: applicable when P
load_2
P
load_1
12 dB.
Three channels in the TCS: applicable when P
load_3
P
load_1
8 dB.
Four channels in the TCS: applicable when:
P
load_2
P
load_1
4 dB.
AND P
load_3
P
load_2
4 dB.
AND P
load_4
P
load_3
4 dB.
When a modem is transmitting at power levels falling outside these loading ranges (which is allowed by the 12 dB
dynamic range), the spurious emissions requirement limits are the power values (in dBmV rather than dBc),
corresponding to the specifications associated with transmissions at the second and higher loaded channels, if the
loading of those channels were decreased (transmit power increased) to just meet the applicability ranges.
With Multiple Transmit Channel mode enabled, the spurious performance requirements do not apply to any upstream
channel from the time the output power on any active upstream channel has varied by more than 3 dB since the last
global reconfiguration time through the end of the next global reconfiguration time changes, excluding transmit
hi
[MULPI]. "Global reconfiguration time" is defined as the inactive
time interval provided between active transmissions which simultaneously satisfies the requirement in Section
B.6.2.20.2 for all TDMA channels in the TCS and the requirement in Section B.6.2.20.3 for all S-CDMA channels in
the TCS. With Multiple Transmit Channel mode enabled the CMTS SHOULD provide global reconfiguration time
to a CM before (or concurrently as) the CM has been commanded to change any active upstream channel transmit
11/17/11 CableLabs
133
power by 3 dB cumulative since its last global reconfiguration time, excluding transmit power changes due
to UCD-induced change in P
hi
[MULPI].
71

dBmV, 45 dBmV, and 45 dBmV. The 41 dBmV power levels are raised for the purposes of determining the spurious
emissions limits, up to 53 dBmV 8 = 45 dBmV.
As an example with four S-CDMA channels in the TCS, at transmit powers per channel of 53 dBmV, 41 dBmV, 41
dBmV, 49 dBmV, 45 dBmV, and 41 dBmV. The 41 dBmV power levels are raised for the purposes of determining
the spurious emissions limits, up to 53 dBmV 4 dB = 49 dBmV; and the artificial 49 dBmV 4 dB = 45 dBmV;
and the artificial 45 dBmV 4 dB= 41 dBmV.
In Table B10, inband spurious emissions include noise, carrier leakage, clock lines, synthesizer spurious products,
and other undesired transmitter products. It does not include ISI. The measurement bandwidth for Inband spurious is
equal to the modulation rate (e.g., 1280 to 5120 kHz). All requirements expressed in dBc are relative to the actual
transmit power that the CM emits in one channel.
72

The measurement bandwidth is 160 kHz for the Between Bursts (none of the channels in the TCS are bursting) specs
of Table B10, except where called out as 4 MHz or 250 kHz.
The Transmitting Burst specs apply during the mini-slots granted to the CM (when the CM uses all or a portion of
the grant), and for 32 modulation intervals before and after the granted mini-slots. The Between Bursts specs apply
except during a used grant of mini-slots on any active channel for the CM, and the 32 modulation intervals before
and after the used grant.
In TDMA mode, a mini-slot may be as short as 32 modulation intervals, or 6.25 s at the 5.12 MHz rate, or as short
as 25 s at the 1.28 MHz rate.
73

Table B10 - Spurious Emissions
74

Inband -40 dBc -72 dBc
Adjacent Band See Table B11 and
Table B12.
-72 dBc
Within the upstream operating range of 5-65 MHz
or 5-85 MHz (excluding assigned channel, adjacent
channels, and carrier-related channels)
See Table B13 and
Table B14.
-72 dBc
For the case where the upstream operating range is
5-65 MHz:
CM Integrated Spurious Emission Limits (all in 250
kHz, includes discretes)
87.5 to 108 MHz

-30 dBmV

-59 dBmV

71
72
73
Paragraph modified per PHYv3.0-N-007.0465-1, #6 on 7/2/07 by KN.
74
Table modified per PHYv3.0-N-11.021-1 and PHYv3.0-N-11.1020-2 on 11/4/11 by PO.
134 CableLabs
11/17/11
5-65 MHz:
CM Integrated Spurious Emissions Limits (all in
4.75 MHz, includes discretes)
1
65 to 87.5 MHz
108 to 136 MHz
2


-40 dBc
-40 dBmV
-45 dBmV

-26 dBmV
-45 dBmV
3
)
5-65 MHz:
1
65 to 87.5 MHz

-50 dBc
-50 dBmV

-36 dBmV
-50 dBmV
5-85 MHz:
CM Integrated Spurious Emissions Limits (all in 4
MHz, includes discretes)
1

85 to 108 MHz
85 to 108 MHz (SHOULD)
108 to 136 MHz

-45 dBc
-50 dBc
-40 dBmV
-45 dBmV

-31 dBmV
-36 dBmV
-40 dBmV
3
)
5-85 MHz:
1

85 to 108 MHz

-50 dBc
-50 dBmV

-36 dBmV
-50 dBmV
Table Notes:
Note 1 These spec limits exclude a single discrete spur related to the tuned received channel; this single discrete
spur MUST NOT be greater than -40 dBmV.
Note 2 The frequencies from 108 to 136 MHz may be forbidden due to national regulations.
Note 3 "dB ref d/s" is relative to the received downstream signal level. Some spurious outputs are proportional to
the receive signal level.

B.6.2.22.1.1 Adjacent Channel Spurious Emissions
Spurious emissions from a transmitted carrier may occur in an adjacent channel which could be occupied by a carrier
of the same or different modulation rate. Table B11 lists the required adjacent channel spurious emission levels for
all combinations of transmitted carrier modulation rates and adjacent channel modulation rates. The measurement is
performed in an adjacent channel interval of appropriate bandwidth and distance from the transmitted carrier, based
on the modulation rates of the transmitted carrier, and the carrier in the adjacent channel.
11/17/11 CableLabs
135
Table B11 - Adjacent Channel Spurious Emissions Requirements Relative to the Per Channel Transmitted
Burst Power Level for Each Channel
interval
Measurement interval and
modulation rate
-50 dBc 160 kHz to 1440 kHz 1280 kHz
-47 dBc 320 kHz to 2880 kHz 2560 kHz
-44 dBc 640 kHz to 5760 kHz 5120 kHz

In addition, the CM SHOULD meet the requirements in Table B12, under the same conditions as listed for Table
B11 above. With a lower power channel bursting as a lone bursting channel (at that point in time), 2 dB relaxation
will apply to the absolute spurious emissions requirement of Table B12.
Table B12 - Adjacent Channel Spurious Emissions Goals Relative to the Per Channel Transmitted Burst
Power Level for Each Channel
Specification in
the interval
Measurement interval and
modulation rate
-51 dBc 160 kHz to 1440 kHz 1280 kHz
-50 dBc 320 kHz to 2880 kHz 2560 kHz
-48 dBc 640 kHz to 5760 kHz 5120 kHz

B.6.2.22.1.2 Spurious Emissions in the Upstream Frequency Range
Table B13 lists the possible modulation rates that could be transmitted in an interval, the required spurious level in
that interval, and the initial measurement interval at which to start measuring the spurious emissions. Measurements
should start at the initial distance and be repeated at increasing distance from the carrier until the upstream band edge
is reached. The lower band edge of the upstream is 5 MHz; the upper band edge of the upstream is 65 MHz or, if
upstream frequency range selection is available in the CM, 85 MHz may be selected.
75

Table B13 - Spurious Emissions Requirements in the Upstream Operating Frequency Range Relative to the
Per Channel Transmitted Burst Power Level for Each Channel
Possible modulation
rate in this interval
Specification in
the interval
1280 kHz -50 dBc 360 kHz to 1640 kHz

75
Paragraph modified by PHYv3.0-N-11.1021-1 11/4/11 by PO.
136 CableLabs
11/17/11
Possible modulation
rate in this interval
Specification in
the interval
2560 kHz -47 dBc 520 kHz to 3080 kHz
5120 kHz -44 dBc 840 kHz to 5960 kHz

In addition, the CM SHOULD meet the requirements in Table B14, under the same conditions as listed for Table
B13 above. With a lower power channel bursting as a lone bursting channel (at that point in time), 2 dB relaxation
will apply to the absolute spurious emissions requirement of Table B14.
Table B14 - Spurious Emissions Goals in the Upstream Operating Frequency Range Relative to the Per
Channel Transmitted Burst Power Level for Each Channel
this interval
Specification in
the interval
Initial measurement interval and distance
from carrier edge
1280 kHz -54 dBc 360 kHz to 1640 kHz
2560 kHz -52 dBc 520 kHz to 3080 kHz
5120 kHz -50 dBc 840 kHz to 5960 kHz

B.6.2.22.2 Spurious Emissions During Burst On/Off Transients
B.6.2.22.3 Modulation Error Ratio (MER)
B.6.2.22.3.1 Definitions
B.6.2.22.3.2 Requirements
Unless otherwise stated, the MER MUST meet or exceed the following limits over the full transmit power range of
Table B9 for each modulation, modulation rate, and over the full carrier frequency range, and for S-CDMA, over
any valid number of active and allocated codes. The 5-65 MHz carrier frequency range refers more precisely to [5
MHz + modulation rate * 1.25 / 2] to [65 MHz - modulation rate * 1.25 / 2]. At the break points between regions, the
higher MER specification applies.
11/17/11 CableLabs
137
Case 1: Flat channel, transmit equalization OFF
76

Case 1a: for modulation rates 2.56 MHz and below for 5 MHz to 65 MHz mode of operation:
MER
symb
MER
symb
MER
symb
Case 1b: for modulation rate 5.12 MHz for 5 MHz to 65 MHz mode of operation:
MER
symb
MER
symb
MER
symb
Case 1c: for modulation rates 2.56 MHz and below for 5 MHz to 85 MHz mode of operation:
MER
symb
MER
symb
MER
symb
Case 1d: for modulation rate 5.12 MHz for 5 MHz to 85 MHz mode of operation:
MER
symb
MER
symb
MER
symb
77

Case 2: Flat channel, transmit equalization ON
Case 2a: for TDMA/QPSK, MER
symb
30 dB.
Case 2b: for S-CDMA and all TDMA modulations except QPSK, MER
symb
35 dB.
Case 2c: for S-CDMA, MER
chip
33 dB.
Case 3: Echo channel, transmit equalization ON
(Echo channel MUST not be limited by dynamic range. Echo channel MUST be a true representation described in
Table B3).
78

Case 3a: In the presence of a single echo selected from the channel micro-reflections defined in Table B3, the
measured MER
symb
MUST be 30 dB for TDMA/QPSK, and 33 dB for S-CDMA and all TDMA modulations,
except QPSK.
Case 3b: In the presence of two or three of the echoes defined in Table B3 (at most one of each specified magnitude
and delay), the measured MER
symb
MUST be 29 dB.
Since the table does not bound echo delay for the -30 dBc case, for testing purposes, it is assumed that the time span
of the echo at this magnitude is less than or equal to 1.5 s.
The CMTS MUST provide a test mode in which it:
Accepts equalizer coefficients via an external interface, (e.g., Ethernet),

76
77
Text added per PHYv3.0-N-11.1021-1 on 11/4/11 by PO.
78
Revised Case 3 per PHYv3.0-N-06.0341-1 by GO on 1/12/07.
138 CableLabs
11/17/11
Sends the coefficients to the CM's pre-equalizer via ranging response message (both set and convolve modes),
Does not adjust the CM's frequency, timing or power.
B.6.2.22.4 Filter Distortion
B.6.2.22.4.1 Amplitude
B.6.2.22.5 Carrier Phase Noise
B.6.2.22.6 Channel Frequency Accuracy
B.6.2.22.7 Modulation Rate Accuracy
B.6.2.22.8 Modulation Timing Jitter
B.6.2.22.8.1 Symbol Timing Jitter for TDMA Upstream Channels in Asynchronous Operation
B.6.2.22.8.2 Modulation Timing Jitter for Synchronous Operation
B.6.2.23 Upstream Demodulator Input Power Characteristics
See Section 6.2.23.
B.6.2.24 Upstream Electrical Output from the CM
The CM MUST output an RF modulated signal with the characteristics delineated in Table B15.
11/17/11 CableLabs
139
Table B15 - Single Channel Electrical Output from CM
79

Parameter Value
Frequency 5 to 65 MHz edge to edge
The following option MAY be provided:
Mode Selection Option:
(Multiple Transmit Channel mode disabled, or
Multiple Transmit Channel mode enabled with
only one channel in the TCS)
TDMA:
P
min
P
min
P
min
to +61 dBmV (QPSK)
S-CDMA:
P
min
where
P
min
P
min
P
min
(two channels in the TCS)
TDMA:
P
min
P
min
P
min
to +58 dBmV (QPSK)
S-CDMA:
P
min
where
P
min
P
min
P
min
(three or four channels in the TCS)
TDMA:
P
min
P
min
P
min
to +55 dBmV (QPSK)
S-CDMA:
P
min
where
P
min
P
min
P
min
Modulation Type QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, and 128-QAM
Modulation Rate (nominal) TDMA: 1280, 2560 and 5120 kHz
S-CDMA: 1280, 2560 and 5120 kHz
Optional pre-3.0 DOCSIS operation:
TDMA: 160, 320, 640 kHz

79
Table updated per PHYV3.0-N-11.1021-1 and PHYv3.0-N-11.1020-2 11/4/11 per PO.
140 CableLabs
11/17/11
Parameter Value
Channel Bandwidth TDMA: 1600, 3200 and 6400 kHz
S-CDMA: 1600, 3200 and 6400 kHz
Optional pre-3.0 DOCSIS operation:
TDMA: 200, 400, 800 kHz
Output impedance 75 ohms
Output Return Loss > 6 dB (across the selected upstream frequency range 5-65 MHz
or 5-85 MHz)
> 6 dB (108 to 862 MHz, or 108 to 1006 MHz if CM center
frequency ranges up to 1002 MHz)
Connector F connector per [ISO/IEC-61169-24] (common with the input)

B.6.2.25 Upstream CM Transmitter Capabilities
80

See Section 6.2.25.
81

B.6.2.25.1 Description of CM Upstream Transmit Channel Set Capabilities
B.6.3 Downstream
B.6.3.1 Downstream Protocol and Interleaving Support
The downstream PMD sublayer MUST conform to [EN 300 429]. The downstream RF output electrical
requirements including downstream frequency plan, interleaver depths, spectrum format, clock, and symbol
requirements are defined in the [DRFI] specification. The CM MUST support the interleaver depths defined in table
A-1 of [DRFI].
B.6.3.2 Downstream Electrical Input to CM
The CM MUST be able to accept any number of signals between one to MDBC simultaneously, located in frequency
intervals that complies with the CMs TB list and Demod List. A CM MUST be capable of being reconfigured to
receive different channels. At a minimum, the CM MUST be able to receive at least 4 downstream channels, located
independently within an arbitrary 64 MHz window in the downstream frequency band. The CM MUST support
bonding of any number of downstream channels up to its maximum. The CM MUST be able to accept RF modulated
signals with the characteristics defined in Table B16.

80
Section modified per PHYv3.0-N-07.0459-2 # 17 on 6/12/07 by KN and per PHYv3.0-N-10.0942-5 on 9/22/10 by JB.
81
Modified per PHY3.0-N-11.1021-1 on 11/4/11 by PO.
11/17/11 CableLabs
141
Table B16 - Electrical Input to CM
82

Parameter Value
Center Frequency 112 to 858 MHz 30 kHz
CM SHOULD tune 112 to 1002 MHz 30 kHz
Level Range (one DOCSIS channel) -17 dBmV to +13 dBmV for 64-QAM
-13 dBmV to +17 dBmV for 256-QAM
Modulation Type 64-QAM and 256-QAM
Symbol Rate (nominal) 6.952 Msym/s (64-QAM and 256-QAM)
Bandwidth 8 MHz (alpha = 0.15 square-root raised cosine shaping for 64-QAM
and 256-QAM)
Total Input Power (80 MHz and above) < 33 dBmV
Maximum average power of carrier input
to CM, within any 8 MHz channel from 78
MHz up to 1006 MHz
X = average power of lowest power demodulated QAM channel

Channels demodulated within the CM
a) min (X + 10 dB, +13 dBmV), for 64-QAM
b) min (X + 10 dB, +17 dBmV), for 256-QAM

8 MHz bands not demodulated within the CM
a) min (X + 10 dB, +20 dBmV), for zero, one, or two 8 MHz bands.
b) min (X + 10 dB, +17 dBmV), for all other 8 MHz bands besides
the two possible allocated exceptions.
Input (load) Impedance 75 ohms
Input Return Loss > 6 dB (across the selected upstream frequency range 5- to 65 MHz
or 5-85 MHz)
> 6 dB (108 to 862 MHz, or 108 to 1006 MHz if CM center
frequency ranges up to 1002 MHz)
Connector F connector per [ISO/IEC-61169-24] (common with the output)

B.6.3.3 CM BER Performance
83

The bit-error-rate performance of a CM is described in this section. For each individual downstream received
channel, the CM MUST meet the BER requirements with the interleaver set to the I = 12, J = 17 mode of
interleaving.
B.6.3.3.1 64-QAM
B.6.3.3.1.1 64-QAM CM BER Performance
-8
when
s
/N
o
) of 25.5 dB or greater. If it is not possible to measure post-FEC BER
directly, Codeword Error Rate, R
C
(as defined below) may be used. In this case, the CM MUST achieve a Codeword
Error Rate of less than or equal to 9 x 10
-7
s
/N
o
) of 25.5 dB or greater.

82
Table modified per ECN-07.0463-1 #1 on 6/12/07 by KN and PHYv3.0-N-11.1020-2 11/4/11 by PO.
83
Paragraph modified per PHYv3.0-N-07.0458 #2 on 6/12/07 by KN.
142 CableLabs
11/17/11
Computation of Codeword Error Rate, R
C
:
( )
( ) ( ) ( )
0 0 0
0
C C E E E E
E E
R
C C U U
U U
C
+ +

=

Where:
E
U
is the value of the count of code words with uncorrectable errors,
E
C
is the value of the count of code words with correctable errors and
C is the value of the count of code words without errors.
Sample the values at the beginning of the test interval (indicated by the subscripted values, E
U0
, E
C0,
and C
0
) and the
end of the test interval (indicated by the un-subscripted values, E
U
, E
C
, and C).
B.6.3.3.1.2 64-QAM Image Rejection Performance
Performance as described in Section B.6.3.3.1.1 MUST be met with analog or digital signal at +10 dBc in any
B.6.3.3.1.3 64-QAM Adjacent Channel Performance
Performance as described in Section B.6.3.3.1.1 MUST be met with a digital signal at 0 dBc in the adjacent
channels.
Performance as described in Section B.6.3.3.1.1 MUST be met with an analog signal at +10 dBc in the adjacent
channels.
Performance as described in Section B.6.3.3.1.1, with an additional 0.2 dB allowance, MUST be met with a digital
B.6.3.3.2 256-QAM
B.6.3.3.2.1 256-QAM CM BER Performance
-8
when
s
/N
o
) as shown below. If it is not possible to measure post-FEC BER directly,
Codeword Error Rate, RC (as defined in Section B.6.3.3.1.1) may be used. In this case, the CM MUST achieve a
Codeword Error Rate of less than or equal to 9 x 10
-7
s
/N
o)
, as shown in
the following:
Input Receive Signal Level E
s
/N
o

-13 dBmV to -6 dBmV 34.5 dB or greater
-6 dBmV to +17 dBmV 31.5 dB or greater
B.6.3.3.2.2 256-QAM Image Rejection Performance
Performance as described in Section B.6.3.3.2.1 MUST be met with an analog or a digital signal at +10 dBc in any
11/17/11 CableLabs
143
B.6.3.3.2.3 256-QAM Adjacent Channel Performance
Performance as described in Section B.6.3.3.2.1 MUST be met with an analog or a digital signal at 0 dBc in the
adjacent channels.
Performance as described in Section B.6.3.3.2.1, with an additional 0.5 dB allowance, MUST be met with an analog
Performance as described in Section B.6.3.3.2.1, with an additional 1.0 dB allowance, MUST be met with a digital
B.6.3.4 Downstream Multiple Receiver Capabilities
See Section 6.3.4.
B.6.3.4.1 Receive Module Parameters
B.6.3.4.2 Receive Channel Parameters
B.6.3.4.3 Standard Receive Channel Profile
In order to limit the complexity in the CMTS of configuring arbitrarily complicated Receive Channel Profiles,
DOCSIS defines a set of "Standard" Receive Channel Profiles that describe a minimal set of constraints.
A CM reports to the CMTS at least one standard RCP as well as its "Manufacturer" RCP that gives more details of
its capabilities and constraints. If the CMTS configures the CM with a Receive Channel Configuration based on a
Standard RCP, some CM capabilities represented in the Manufacturer RCP but outside the Standard RCP may not be
available.
The Standard Receive Channel Profile for the European technology option named "CLAB-8M-004" is defined for 8
MHz operation as depicted in Figure B2. The RF input from the cable plant is connected to a Receive Module with
a block of 8 adjacent channels. This Receive Module represents the constraints of a tuner with a 64 MHz capture
bandwidth that can be placed at any point in the full DOCSIS frequency range. The demodulation of four channels
anywhere within the capture bandwidth is represented by the four Receive Channels, each corresponding to the
packet stream from a single QAM channel output to the MAC layer.
84

84
This paragraph and following figure caption modified by PHY3.0-N-11.1020-2 11/4/11 by PO
144 CableLabs
11/17/11
Tuner
Receive Module 1
[ Adjacent Channels Constraint: 8 ]
RC1 RC2 RC3 RC4
MAC Sublayer
Receive
Channels
(each cor-
responds to
single QAM
channel)
Packets
RF Port
Diplexer
Splitter

Figure B2 - 8 MHz DOCSIS Standard Receive Channel Profile CLAB-8M-004

For this Standard Receive Channel Profile, the CMTS sends the following parameters to the CM in the Receive
Channel Configuration:
Center frequency of the first channel in the 8-channel Receive Module.
Center frequency of each Receive Channel within the capture bandwidth of the Receive Module.
B.6.3.4.4 Example Manufacturer Receive Channel Profile
B.6.3.4.5 Reconfiguration of CM Receive Channels
B.6.3.5 Non-Synchronous DS Channel Support
See Section 6.3.5.
11/17/11 CableLabs
145
Annex C MPEG Header Synchronization and Recovery
The downstream bitstream is defined as a continuous series of 188-byte packets. These packets consist of a 4-byte
header followed by 184 bytes of payload. This transport stream is specified in MPEG-2 Part 1, Systems [ISO 13818]
with modifications described in [ITU-T J.83-B].
C.1 MPEG Header Synchronization and Recovery in the North American
Technology Option
When implementing the first physical layer technology option referred to in Section 1 (1.1 Introduction and Purpose)
and specified in Section 6, modifications described in [ITU-T J.83-B], in the section referring to the MPEG-2
transport framing apply to the transport stream format.
The MPEG-2 packet stream SHOULD be declared "in frame" (i.e., correct packet alignment has been achieved)
when five consecutive correct parity checksums, each 188 bytes from the previous one, have been received.
The MPEG-2 packet stream SHOULD be declared "out of frame", and a search for correct packet alignment started,
when nine consecutive incorrect parity checksums are received.
C.2 MPEG Header Synchronization and Recovery in the European Technology
Option
When implementing the second physical layer technology option referred to in Section 1 (1.1 Introduction and
Purpose) and specified in Annex B, modifications described in EN 300 429 [EN 300 429] apply to the transport
stream format.
The MPEG-2 packet stream SHOULD be declared "in frame" (i.e., correct packet alignment has been achieved)
when five consecutive correct sync bytes, each 188 bytes from the previous one, have been received.
The MPEG-2 packet stream SHOULD be declared "out of frame", and a search for correct packet alignment started,
when nine consecutive incorrect sync bytes are received.
146 CableLabs
11/17/11
Appendix I Example Preamble Sequence
I.1 Introduction
A programmable preamble superstring, up to 1536 bits long, is part of the channel-wide profile or attributes,
common to all burst profiles on the channel, but with each burst profile able to specify the start location within this
sequence of bits and the length of the preamble. The first bit of the Preamble Pattern is designated by the Preamble
Value Offset. The first bit of the Preamble Pattern is the first bit into the symbol mapper (Figure 62 and Figure 6
3), and is the first symbol of the burst (see Section 6.2.13). As an example, for Preamble Offset Value = 100, the
101st bit of the preamble superstring is the first bit into the symbol mapper, and the 102nd bit is the second bit into
the mapper, and is mapped to Q1, and so on. An example 1536-bit-long preamble superstring is given in Appendix
I.2.
I.2 Example Preamble Sequence
The following is the example1536-bit preamble sequence:
Bits 1 through 128:
1100 0011 1111 0000 0011 0011 1111 1100 0011 0011 0000 0011 1100 0000 0011 0000
0000 1110 1101 0001 0001 1110 1110 0101 0010 0101 0010 0101 1110 1110 0010 1110
Bits 129 through 256:
0010 1110 1110 0010 0010 1110 1110 1110 1110 1110 0010 0010 0010 1110 1110 0010
1110 1110 1110 0010 1110 0010 1110 0010 0010 0010 0010 1110 0010 0010 1110 0010
0010 1010 0110 0110 0110 1110 1110 1110 0010 1110 0010 1110 0010 1110 0110 1010
0010 1110 1110 1010 0110 1110 0110 0010 0110 1110 1010 1110 0010 1010 0110 0010
0010 1110 0110 1110 0010 1010 1010 0110 0010 1110 0110 0110 1110 0010 0010 0110
0010 1110 0010 1010 0010 1110 0110 0010 0010 1010 0010 0110 0010 1010 0010 1010
0010 1110 0110 1110 0110 0110 1110 0010 0110 1010 0110 0010 1110 1110 1010 0010
1110 1110 0010 1110 1110 1110 0010 1110 1110 0010 1110 0010 0010 1110 0010 0010
1110 1110 1110 0010 0010 0010 1110 0010 1110 1110 1110 1110 0010 0010 1110 0010
1110 0010 0010 0010 1110 1110 0010 0010 0010 0010 1110 0010 0010 0010 0010 1110
0011 0000 1111 1100 0000 1100 1111 1111 0000 1100 1100 0000 1111 0000 0000 1100
0000 0000 1111 1111 1111 0011 0011 0011 1100 0011 1100 1111 1100 1111 0011 0000
1100 0011 1111 0000 0011 0011 1111 1100 0011 0011 0000 0011 1100 0000 0011 0000
0000 1110 1101 0001 0001 1110 1110 0101 0010 0101 0010 0101 1110 1110 0010 1110
0010 1110 1110 0010 0010 1110 1110 1110 1110 1110 0010 0010 0010 1110 1110 0010
1110 1110 1110 0010 1110 0010 1110 0010 0010 0010 0010 1110 0010 0010 1110 0010
11/17/11 CableLabs
147
0010 0010 1110 1110 1110 1110 1110 1110 0010 1110 0010 1110 0010 1110 1110 0010
0010 1110 1110 0010 1110 1110 1110 0010 1110 1110 0010 1110 0010 0010 1110 0010
Bits 1281 through 1408
1100 1100 1111 0000 1111 1111 1100 0000 1111 0011 1111 0011 0011 0000 0000 1100
0011 0000 0011 1111 1111 1100 1100 1100 1111 0000 1111 0011 1111 0011 1100 1100
0011 0000 1111 1100 0000 1100 1111 1111 0000 1100 1100 0000 1111 0000 0000 1100
0000 0000 1111 1111 1111 0011 0011 0011 1100 0011 1100 1111 1100 1111 0011 0000
148 CableLabs
11/17/11
Appendix II S-CDMA Framing
This appendix is informative. In case of conflict between this section and any normative section of this specification,
the normative section takes precedence.
Please note that the pseudo-code in Sections II.1 and II.2 below is specific to the case of a single burst using all
spreading codes.
II.1 Coded Subsymbol Numbering
The following code sample contains a short algorithmic description of the operation of the address generator for the
coded subsymbols. The address generator for the coded subsymbols fills rows first using the interleaver step size
parameter (step in the listing) to step through the spreading intervals within a row. Each step is performed using a
modified modulo algorithm which allows the use of interleaver step size and spreading intervals per frame with
common divisors. After each row is filled, the next row is begun with the first spreading interval. In the following
listings, the index "i" is initialize to the value "1" and coded_col0 is defined as "0".
for(row = FIRST_ROW; row LAST_ROW; row++)
{
coded_col = 0;
store_coded( row, coded_col, cod_sym);
/* Store the coded portion of the symbol (or preamble) to (row, coded_col) */
for( i = 1; i < framelen; i++ )
{
coded_col = coded_col + Interleaver_step_size;

if( mod( i, framelen / gcd( step, framelen ) ) == 0 )
coded_col = coded_col + 1; /* gcd is greatest common divisor */

coded_col = mod( coded_col, framelen );
store_coded( row, coded_col, cod_sym );
/* Store the coded portion of the symbol (or preamble) to (row,
coded_col) */
}
}
II.2 Uncoded Subsymbol Numbering
The following is a short algorithmic description of the operation of the address generator for uncoded subsymbols.
The generator fills columns within a subframe first. The row index increments by one for each uncoded subsymbol.
At the end of the subframe, the column index is incremented and the row index set to the first row of the subframe.
After completing a subframe, the next subframe begins with the next uncoded subsymbol.
uncoded_col = 0;
uncoded_row = FIRST_ROW;
while( uncoded_row LAST_ROW)
{
if( ( uncoded_row + R ) > LAST_ROW )
Rprime = LAST_ROW - uncoded_row + 1;
else
Rprime = R;

for( i = 0; i < Rprime; i++)
{
/* Check whether (uncoded_row, uncoded_col) is a preamble location.
* If it is, go to next location */
if( not_preamble( uncoded_row, uncoded_col ) )
store_uncoded( uncoded_row, uncoded_col, unc_sym );
11/17/11 CableLabs
149
uncoded_row = uncoded_row + 1;
}

uncoded_row = uncoded_row - Rprime;
uncoded_col = uncoded_col + 1;
if (uncoded_col framelen)
{
uncoded_col = 0;
uncoded_row = uncoded_row + R;
}
}
FIRST_ROW and LAST_ROW are, respectively, the first and last row (i.e., code) in each frame spanned by the
grant. FIRST_ROW can be between 0 and 127 in the first frame of the allocation and is 0 in any other frames that
the grant may span (if any). LAST_ROW can be between 0 to 127 in the last frame of the burst and is 127 for any
other frame (if any).
II.3 Framer Output Numbering
The following code sample contains a short algorithmic description of the operation of the address generator for the
output symbols. The address generator for the output symbols is used to access both the coded and uncoded
subsymbol memories. The output address generator accesses all of the rows (codes) of a spreading interval first
followed by subsequent spreading intervals. This algorithmic description is generalized to any allowable number of
active codes, where the Number of Active Codes parameter is given by N
a
.
for( col=0; col < framelen; col++ )
for( row=(128-Na); row < 128; row++ )
outsym = get_data( row, col );

Note that the Framer row addressing operates independently of the Selectable Active Codes mode. In either
Selectable Active Codes Mode 1 or 2, the unused codes are at the lower row address indices of the frame and,
therefore, are not used.
II.4 Comments
In the preceding code samples, the number of iterations for the loop is not always correct since an allocation can be
less than the number of codes. In II.2, the listing supports the case of a shortened sub-frame.
150 CableLabs
11/17/11
Appendix III Ambient Temperature and Wind Loading Effects
This appendix discusses possible ambient temperature change and dynamic wind loading effects relevant to
operating a system with DOCSIS 2.0 CMs and CMTSs. The intent of this appendix is to describe possible
approaches for dealing with these issues. The relationships between the timing variation of the received upstream
signal and the rate of change of these ambient plant conditions are discussed. However, measured field data
providing the statistics of the ambient conditions used in these relationships is not available, so it is not possible at
the time of writing to determine the magnitude or frequency of occurrence of these conditions on operational cable
systems. This appendix is not intended to be an exhaustive discussion of either these issues or solutions.
The following issues are discussed in this appendix:
Synchronization tolerances to plant delay variations;
Changes in propagation delay due to temperature changes;
Changes in propagation delay due to wind in the case of aerial cable plant.
III.1 Synchronization Tolerances to Plant Delay Variations
The CMTS receiver synchronization requirements for S-CDMA and Advanced TDMA are identical for the same
signal constellation and symbol rates. However, for S-CDMA, burst synchronization is accomplished, to a fine
degree, through the ranging process, while for TDMA, burst synchronization is accomplished to a coarse degree,
through the ranging process, and then to a fine degree through a receiver burst timing recovery process. In both
cases, the degree of timing accuracy required in the receiver is tighter for higher symbol rates and higher-order
constellations.
Because S-CDMA requires a fine degree of timing accuracy to be accomplished solely by the ranging process, it is
more sensitive to changes in the propagation delay of the cable plant between ranging intervals, which can be as
much as 30 s apart. Table III1 lists plant delay variations that can be accommodated in S-CDMA and TDMA
modes for a 1 dB degradation under example conditions.
Table III1 - Allowable Plant Timing Drift
Constellation E
s
/N
o
for 1e-8
BER(dB)
Allowable peak-peak
plant delay variation
(ns)S-CDMA mode
Allowable peak-peak
plant delay variation
(ns)TDMA mode
Fully-coded QPSK 5 90 800
TCM QPSK 9 79 N/A
TCM 8-QAM 12 57 N/A
Uncoded QPSK 15 38 800
Fully-coded 64-QAM 17.7 24 800
TCM 32-QAM 19 18 N/A
Uncoded 16-QAM 22 9 800
TCM 128-QAM 25 6 N/A

11/17/11 CableLabs
151
Defined conditions:
1 dB degradation at 1e-8 BER;
Uniform ranging offset over 1/64 chip;
63 CMs, each with 2 codes;
E
s
/N
o
numbers are ideal theoretical values, not including implementation effects;
5.12 MHz modulation rate;
Timing variation over 30 s period;
TDMA receiver accepts 2 symbol coarse timing offset (implementation-dependent).
This channel impairment should be considered along with all of the other upstream channel characteristics
highlighted in Table 52 or in Table B3, depending on which of the two technology options is relevant in the
particular application (North American or European) as described in Section 1.1.
DOCSIS requires station maintenance at least every 30 s (T4 time out has a maximum value of 35 s). For S-CDMA
at a given modulation and symbol rate, if there exists a rapid propagation delay variation such that the resulting delay
change cannot be tracked by station maintenance, then one or more of the following performance trade-offs and/or
system changes may be enacted: (1) decrease the station maintenance period; (2) decrease the constellation order; (3)
decrease the modulation rate; (4) apply additional error correction; (5) apply some combination of 1 through 4, or (6)
change the channel to operate in TDMA mode.
The following sections discuss the relationship of temperature changes and wind loading on the propagation delay in
coaxial and HFC cable plants.
III.2 Change in Propagation Delay Due to Temperature Changes
III.2.1 Fiber Delay Changes Due to Temperature
In HFC plant design, the number of amplifiers in cascade in the coax portion is kept low in order to keep signal
degradation to an acceptable level. As a result, long runs of cable plant are mainly comprised of fiber. A typical
value for propagation delay variation due to temperature change of the fiber is 44 ps per km per degree C. The delay
variation comes mostly from the change in refractive index of the glass with temperature, not the change in fiber
length.
It is assumed that changes in optical cable length due to stretching or expansion will be a negligible factor because
optical cables are built to isolate the fiber from stresses in the cable itself. The fiber usually sits loosely in a tube
inside the cable and quite a bit of relative movement is possible. This construction allows normal cable handling and
aerial deployment without resulting in high stress on the optical fiber.
Assuming 44 ps per km per degree C, any product of optical cable length and temperature change which equals 50,
results in approximately a 2 ns change in the fiber propagation delay. For example, 25 km fiber and a 2 degree
temperature change will result in a 2 ns change in propagation delay. For the maximum distance between CMTS and
CM specified in DOCSIS of 100 miles or roughly 160 km, the temperature change needed for a 2 ns change in one-
way propagation delay is 0.3 degree C.
152 CableLabs
11/17/11
Obviously, the issue is how fast the cable core (where the fiber is) will heat up under ambient temperature changes.
For buried or underground cable, there is no issue. For aerial cable solar loading changes should be considered.
Black sheathed cable has interior temperatures quite a bit higher than ambient in sunlight, but data is currently
unavailable. When the rising sun hits aerial cable on a cold morning, one would expect a temperature change.
Similarly, sunlight appearing out of cloud cover may have a similar impact although the size of the shadow of the
cloud moving out of the way has to be considered. The numerical examples above suggest that only long aerial cable
runs may have a problem under some combinations of time-of-day and weather.
III.2.2 Coaxial Cable Delay Changes Due to Temperature
The coaxial cable has "blown" foam between the center conductor and the solid shield. The nominal propagation
velocity is about 87% of free space velocity. Propagation velocity does not vary markedly with temperature. Given
the relatively short lengths of coaxial cable in most HFC plants (a few miles), this seems unlikely to be a significant
source of delay variation.
III.2.3 Delay Change Due to Wind
Aerial cable stretches during windy conditions, so it is possible to estimate a propagation delay from the change in
length under various wind loads. As mentioned, the construction of optical cable makes it tolerant of stretching, so it
is assumed that optical cable stretching, due to wind loading, can be ignored. Wind loading will affect aerial coaxial
cables.
Wind loading is difficult to deal with, analytically, because it is unlikely to be uniform along the cable. A delay
model derived from a significant body of measured data is needed to investigate this further. Wind loading may be a
source of fast delay variation and station maintenance may not occur at intervals small enough for the ranging
mechanism to track this variation accurately.
The effect of wind loading on a typical cable was investigated with a publicly available program from a coaxial cable
manufacturer. These calculations showed that length changes in the range 0.01% and 0.05% are possible for various
amounts of wind loading. This converts to significant propagation delay variation. For example, at 5 miles (8 km)
and 0.02% length variation, the change in propagation delay is:
(8/3e5)*(1/0.87)*2e-4 s = 6 ns
This is a peak value, but the length of coax is quite short and the wind load is moderate. While the time duration over
which this delay variation occurs is unspecified, it may be noted that wind gust data is readily available for most
cities, and wind gust will be the primary mechanism for wind-based timing changes on cable plants. For example, in
New York City at the time of this writing, wind gusts of up to 40 mph were reported, while average wind speed was
about 10 mph. Hence, over a period of 1 to 4 s (the typical wind gust measurement interval), the wind speed changed
by 30 mph. Much stronger wind gusts are frequently measured in locations prone to windy conditions.
11/17/11 CableLabs
153
Appendix IV Description of Upstream Transmit Channel Set Capability:
Example Calculations for Reporting and Figuring the Number
of Active Channels Supported
As described in the requirements in Section 6.2.25.1, the CM communicates to the CMTS its capabilities in regards
to the number of active upstream channels it can support. The method of conveyance must clearly and
unambiguously describe the capabilities of the CM to the CMTS.
This appendix provides the definition for the parameters as in Section 6.2.25.1, and also provides examples
illustrating the calculations that need to take place at both the CM and the CMTS so that precisely the same possible
upstream transmit channel combinations are understood at both.
Let X = number of 6.4 MHz channels, Y = number of 3.2 MHz channels, and Z = number of 1.6 MHz channels that
are active at one time (the Transmit Channel Set). Let X
max
be the maximum number of 6.4 MHz channels that the
CM can support, Y
max
be the maximum number of 3.2 MHz channels that the CM can support, and Z
max
be the
maximum number of 1.6 MHz channels that the CM can support. Z
max
is equivalent to the maximum number of
upstream transmitters that the CM can support.
The CM reports X
max
, Y
max
, and Z
max
during registration using modem capabilities encodings specified in [MULPI],
in the section entitled, "Modem Capabilities Encoding." It is specified in Section 6.2.25.1 that X
max
Y
max
Z
max
.
Communications of X
max
, Y
max
, and Z
max
by a CM provides enough information to describe all the combinations
possible for X, Y, and Z.
It is understood that these then directly provide the combinations of (X, Y, Z) which are (X
max
, 0, 0); (0, Y
max
, 0);
and (0, 0, Z
max
).
From these values the CMTS may calculate B
max
= max (6.4*X
max
, 3.2*Y
max
, 1.6*Z
max
).
Then, all values for (X, Y, Z) are allowed which satisfy all inequalities below:
6.4*X + 3.2*Y + 1.6*Z B
max

X X
max

Y Y
max

X + Y + Z Z
max
.
Examples:
EX1.
X
max
= 4; Y
max
= 4; Z
max
= 4.
In this case all combinations with X + Y + Z 4 are possible.
EX2.
X
max
= 4; Y
max
= 8; Z
max
=16.
154 CableLabs
11/17/11
In this case we have all combinations where 6.4X + 3.2Y + 1.6Z 25.6.
This is identical to a receiver which can channelize 25.6 MHz into any combination of the three channel bandwidths
without restriction.
EX3.
X
max
= 4; Y
max
=8; Z
max
=8.
This is a subset of Example 2.
With this capability reporting, both conditions a) 6.4X + 3.2Y + 1.6Z 25.6, and b) X + Y + Z 8, must be met.
Thus, all cases in Example 2 with X + Y + Z > 8 ARE EXCLUDED with this reporting.
The CM in this case only has to support (and will only be asked to support) a maximum of 8 channels, even if they
are all 1.6 MHz bandwidth.
There are no other restrictions on the Transmit Channel Set which must be supported.
EX4.
X
max
= 4; Y
max
= 6; Z
max
= 8.
This is a subset of Example 2 and a subset of Example 3.
With this capability reporting, both conditions a) 6.4X + 3.2Y + 1.6Z 25.6, and b) X + Y + Z 8, must be met, but
in addition, Y 6.
Thus, all cases in Example 2 with X + Y + Z > 8 ARE EXCLUDED with this reporting, and ALSO EXCLUDED are
(0, 8, 0); (0, 7, 1); {and (0, 7, 0)}.
The latter cases are EXCLUDED because the cases with Y > Y
max
= 6 have to be excluded.
Note that (1, 6, 0); (0, 6, 2); {and (0, 6, 1) and (0, 6, 0)} are INCLUDED.
11/17/11 CableLabs
155
Appendix V Description of Upstream Channel Power Control with
Multiple Upstream Channels
This appendix provides an explanation of the difference in the power control algorithm with a) multiple upstream
channels, compared to b) previous DOCSIS operation. This appendix also explains the extension of some DOCSIS
2.0 parameters and defines new parameters in the upstream power control operation, and also provides examples
illustrating the upstream power adjustment and control by the CMTS of CM's which have Multiple Transmit Channel
mode enabled.
V.1 DOCSIS 2.0 Parameters Extended to Multiple Transmit Channel mode
The basic parameters from DOCSIS 2.0 such as P
r
, P
hi
, P
low
, and for S-CDMA with Maximum Scheduled Codes
Enabled, P
hi_T
, P
low_S
, and P
on
, maintain their definitions and roles from DOCSIS 2.0. However, these are extended
with Multiple Transmit Channel mode enabled so that they are defined for each upstream channel which the CMTS
has activated for the CM. These parameters are denoted with a subscript identifying the channel for which they are
associated. Therefore, with four upstream channels activated by the CMTS for a particular CM, that CM will have
P
r_1
, P
r_2
, P
r_3
, and P
r_4
for its reported power level (dBmV) for its four upstream channels, respectively.
V.2 New Parameters in DOCSIS 3.0 Upstream Power Control
("loading," P
load_n
, P
load_min_set
, Dynamic Range Window, P
low_multi_n
)
85

In addition to the extension of these DOCSIS 2.0 parameters, there are new parameters introduced in DOCSIS 3.0
upstream power control. Parameters such as P
load_min_set
, P
load
_
n
(e.g., P
load_1
, P
load_2
, etc.), and P
low_multi n
are added, and
new concepts such as Dynamic Range Window and "loading" an upstream digital-to-analog converter with the
modulated waveforms to be transmitted upstream are added.
86

The "loading" of a modulated waveform to be transmitted upstream is a measure of how close to the maximum
power level the waveform happens to be (in terms of its average power); this is dependent on the absolute power
level of the waveform and of the modulation type of the waveform, and also the number of channels in the Transmit
Channel Set. For example, with 64-QAM TDMA modulation the maximum power level allowed is 57 dBmV, 54
dBmV, and 51 dBmV for one, two, and three channels, respectively, in the Transmit Channel Set. A "fully loaded"
64-QAM TDMA waveform with one channel in the Transmit Channel Set would be transmitted at 57 dBmV while a
"fully loaded" 64-QAM waveform with three channels in the Transmit Channel Set would be transmitted at 51
dBmV. A 64-QAM TDMA waveform which is "underloaded" by 15 dB would transmit at 57 dBmV - 15 dB = 42
dBmV with one channel in the Transmit Channel Set, and a 64-QAM TDMA waveform which is "underloaded" by
15 dB would transmit at 51 dBmV - 15 dB = 36 dBmV with three channels in the Transmit Channel Set. The
parameter P
load_n
is used to convey the amount by which the nth channel is underloaded. Thus, in the case of the
single channel in the Transmit Channel Set, with 42 dBmV for the 64-QAM TDMA transmit level, the channel is
underloaded by 15 dB, therefore, P
load_1
= 15 dB. Similarly, in the case of three channels in the Transmit Channel
Set, with three 64-QAM TDMA modulations transmitting at 41 dBmV, 38 dBmV, and 36 dBmV, we have P
load_1
=
10 dB, P
load_2
= 13 dB, and P
load_3
= 15 dB, where P
load_1
corresponds to the channel with the lowest value of P
load
, or
equivalently, the highest or more full loading.
The concept of "loading," and the values of P
load_n
, are only indirectly tied to the absolute transmit power, as seen in
the preceding paragraph, with the absolute transmit power equaling the maximum transmit power (for the modulation
and number of channels in the Transmit Channel Set) minus P
load_n
for the n
th
channel.

85
Revised this title per PHYv3.0-N-06.0340-1 by GO on 1/12/07.
86
156 CableLabs
11/17/11
For the case with multiple modulation types across the burst profiles of a channel, such as QPSK TDMA and 64-
QAM TDMA burst profiles existing in one channel, the maximum power for the channel is determined by the lower-
valued 64-QAM level of 57 dBmV compared to QPSK's 61 dBmV (assuming for this example a single channel in
the Transmit Channel Set), just as in DOCSIS 2.0. P
hi_1
= 57 dBmV for this case, as set by the 64-QAM TDMA
maximum power. With "full loading" on this channel, 64-QAM TDMA will transmit at 57 dBmV. Note that QPSK
will transmit at 57 dBmV - 1.2 dB = 55.8 dBmV in this situation (due to the constellation gain of -1.18 dB for
QPSK). While the 64-QAM TDMA modulation is "fully loaded," and P
load_1
= 0 dB in this example, the QPSK
TDMA modulation on this channel is underloaded, in fact, by 5.2 dB. P
hi_1
is determined by the 64-QAM modulation
in this case, and the QPSK modulation is NOT determining P
hi
(by virtue of the definition of P
hi
= min(P
max
- G
const
)
over all burst profiles used by the CM in this channel). Therefore, it is only the loading of the 64-QAM modulation
which is described by P
load_1
. The loading occurring with QPSK in this example is irrelevant in the calculations and
parameters involved in power control with Multiple Transmit Channel mode enabled.
The parameter P
load_min_set
is commanded to the CM by the CMTS, and sets in place the restriction on P
load_n
that
P
load_min_set
P
load_n
P
load_min_set
+ 12 dB, for all channels in the Transmit Channel Set.
Since P
load_n
= P
hi_n
- P
r_n
, the value of P
load_min_set
places the limit on P
r_n
for each channel that
P
hi_n
- P
load_min_set
P
r_n
P
hi_n
- P
load_min_set
- 12 dB.
(Note that P
low_multi_n
= P
hi_n
- P
load_min_set
- 12 dB.)
87

Therefore, for each channel in the Transmit Channel Set, the range of P
r_n
for that channel is restricted to operating
over a window of 12 dB by the value commanded by the CMTS. This is the Dynamic Range Window for the CM for
the n
th
channel. Note that the actual transmit values, P
r_n
, for each channel in the Transmit Channel Set may operate
over a DIFFERENT 12 dB range, but the P
load_n
values must all fall within the SAME 12 dB window. Therefore,
there IS a unique Dynamic Range Window set for the CM by P
load_min_set
, however the absolute transmit range for
each channel may differ. For example, let there be two channels in the Transmit Channel Set, and one channel has
only 64-QAM TDMA modulation and the other has only QPSK TDMA modulation, and further, P
load_min_set
= 15 dB.
For each channel the Dynamic Range Window is such that 15 dB P
load_n
27 dB. However, the channel with only
64-QAM TDMA modulation will operate with P
r_n
between 54 dBmV - 15 dB = 39 dBmV and 54 dBmV - 15 dB -
12 dB = 27 dBmV, while the channel with only QPSK TDMA modulation will operate with P
r_n
between 59.2 dBmV
- 15 dB = 44.2 dBmV and 59.2 dBmV - 15 dB - 12 dB = 32.2 dBmV.
V.3 Example Upstream Power Control with Multiple Transmit Channel mode Enabled
In DOCSIS 2.0, when a change to the power level is commanded to the CM, the new P
r
is tested against P
hi
, and P
low
,
and P
r
is clipped at P
hi
or P
low
if in fact P
r
would have been higher than P
hi
or lower than P
low
(per [DOCSIS RFI
2.0]). With MSC enabled, P
r
is tested and clipped against P
hi_T
and P
low_S
, and P
on
is tested and clipped against P
hi_S
.
When a CM is placed in Multiple Transmit Channel mode by the CMTS, P
load_min_set
must be provided to the CM.
This will set the Dynamic Range Window for the CM.
With MTC mode enabled, each channel continues operating with the testing and clipping as described for DOCSIS
2.0, with the following exceptions:
1. An additional test is performed comparing the newly proposed P
r_n
with P
hi_n
- P
load_min_set
, to restrict the
maximum power level for the channel to be P
load_min_set
below P
hi_n
. Effectively, the highest reported power level
for each channel is reduced by P
load_min_set
(dB).

87
Revised this statement per PHYv3.0-N-06.0340-1 by GO on 1/12/07.
11/17/11 CableLabs
157
2. If when comparing the newly proposed P
r_n
against P
hi_n
- P
load_min_set
, it is found that P
r_n
is too high, then the
power change command for P
r_n
is to be ignored. (Note that when P
load_min_set
is 0 dB there will not be discarding
of commands for going over the power range limit; this is because the protocol described in Section 6.2.19
provides that P
r_n
is first tested and clipped against P
hi_n
, without discarding if clipping is needed.)
3. The steps of #1 and #2 are performed for checking for the low power levels, too, where now P
r_n
is not allowed
to drop below P
hi_n
- (P
load_min_set
+ 12 dB).
4. For MSC enabled, P
r_n
is tested to ensure it is not exceeding P
hi_T_n
- P
load_min_set
, and also that P
r_n
is not below
P
hi_T_n
- (P
load_min_set
+ 12 dB), and if either occurs the command changing P
r_n
is ignored.
88

5. For MSC enabled, P
on
is clipped so that it will not exceed P
hi_S
- P
load_min_set
.
Consider an example with four TDMA channels, each with 64-QAM only. In this example, suppose the CMTS
begins with P
load_min_set
= 3 dB; then for each channel the channel power, P
r_n
, is limited to be <= P
hi_n
- 3 dB, so that
each channel is limited to 48 dBmV rather than the four-channel max for 64-QAM which is 51 dBmV. The lowest
value allowed for each P
r_n
is 36 dBmV (12 dB below the highest value.) Consider that each channel is set for P
r_n
=
46 dBmV. This allows 2 dB of headroom before P
load_min_set
would have to be increased to allow P
r_n
to rise above 48
dBmV.
In this example, consider that the CMTS then commands one of the channels to drop its power level by 2 dB. This
change would require a reconfiguration time on that channel, but all other channels may continue to receive grants
and be bursting, and all fidelity requirements would have to still be met.
Then consider that the CMTS commands this channel to drop its power another 2 dB, and provides that channel a
reconfiguration time. Now this channel has had its power changed by 4 dB since the last global reconfiguration time
was provided, so the fidelity requirements are not enforced.
Now consider that this channel is moved back up by 2 dB, so its net change since the last global reconfiguration time
is back to -2 dB. The fidelity requirements are still not in place; they were removed when the CM was commanded to
move one of its channels by more than 3 dB since the last global reconfiguration time, and the specification states
that the fidelity requirements are not back in place until the next global reconfiguration time.
To continue the example, assume again that the channel is dropped another 2 dB, down once again to 42 dBmV, 4
dB below the starting value.
When a global reconfiguration time is subsequently provided for the CM, all conditions are met for full fidelity
requirements to be applicable.
Then, this channel is commanded to drop its power another 2 dB, and again it is provided a reconfiguration time
such that it can implement this change; no global reconfiguration time is needed or provided. Now the fidelity
requirements are slightly relaxed, because this channel loading is 6 dB higher (i.e., lower power) than the next
closest channel loading (40 dBmV versus 46 dBmV). (Also note that it is the channel loading, i.e., P
r_n
compared to
P
hi_n
, and NOT the absolute value of P
r_n
, that determines if the 4 dB range for 4 active channels is met.)
Continuing this example further, consider that another channel has its power dropped by 4 dB, and a reconfiguration
time is provided on this channel, but until a global reconfiguration time is provided the fidelity requirements are not
enforced (due to this 4 dB change since the previous global reconfiguration). Once a global reconfiguration time is
provided, the full fidelity requirements will be in place again, since all conditions are met (the P
load_n
values are 5 dB,
5 dB, 9 dB, and 11 dB and the lowest P
load_n
value is within 3 dB of P
load_min_set
which is still 3 dB).
Now consider that the CMTS wishes to raise the power of one of the two remaining high-power channels by 2.5 dB.
Unless P
load_min_set
is reduced by at least 0.5 dB, to 2.5 dB or less from its current value of 3 dB, this power change
command would be ignored since it would result in a violation of the Dynamic Range Window for the channel. Let

88
Revised this step per PHYv3.0-N-06.0340-1 by GO on 1/12/07.
158 CableLabs
11/17/11
the CMTS send a change in P
load_min_set
to 1 dB, and at the same time increase the power in one of the two remaining
channels by 2.5 dB. The CM will need to wait for a global reconfiguration time to make the dual changes in
P
load_min_set
and in P
r_n
for the channel moving up to 48.5 dBmV. If all these changes are implemented at the CM, then
for each channel, since each has only 64-QAM TDMA, the maximum value of P
r_n
(the top of the Dynamic Range
Window) is 50 dBmV, and the four channels are commanded to 48.5 dBmV, 46 dBmV, 42 dBmV, and 40 dBmV. If
the CMTS provides a global reconfiguration time, then these changes can be implemented. However, if the CMTS
provides reconfiguration time to the one commanded channel, but this is not a global reconfiguration time, then the
CM will not be able to implement the command to increase its power, since global reconfiguration time is required
for the change in the Dynamic Range Window. Note that even if the pending power change did not require the new
P
load_min_set
, since the channel power change was commanded concurrently with the change in the Dynamic Range
Window, the CM should not implement the channel power change until the next global reconfiguration time. The
CM should wait to implement both changes during the next global reconfiguration time, at which time the CM will
still be meeting all fidelity requirements. Note that the P
load_n
values are now 2.5 dB, 5 dB, 9 dB, and 11 dB.
As a further refinement of this example, consider that the channel at 40 dBmV actually had a QPSK TDMA
modulation in its burst profiles, along with the 64-QAM. Let this channel also be commanded to reduce its power by
another 2 dB, to P
r_n
= 38 dBmV. P
load_n
for this channel is thus 51 dBmV - 38 dBmV = 13 dB, which is exactly 12
dB below P
load_min_set
, and thus P
r_n
is exactly equal to P
low_multi_n
= 51 dBmV - 1 dB - 12 dB = 38 dBmV in this
example. There is no other change in this example. However, note that when QPSK is bursted on this channel it is
bursting at 38 dBmV - 1.18 dB = 36.8 dBmV. Note that this is BELOW the power level of P
low_multi
_
n
for this
channel. This is loading, compared to QPSK TDMA with four channels, which is 55 dBmV - 36.8 dBmV = 18.2 dB,
while P
load_min_set
is 1 dB. Even though QPSK on this channel is underloaded by 18.2 dB, and this is 17.2 dB from
P
load_min_set
, the P
load
for this channel is still 12 dB. The fact that the QPSK modulation is more than 17 dB
underloaded compared to P
load_min_set
is not relevant. The CM must be able to accommodate transmission of the
QPSK modulation at the slightly less loaded value than occurs with 64-QAM, since the power of the QPSK bursts
compared to the 64-QAM bursts is only differing by 1.18 dB; the fact that the QPSK bursts are more underloaded
(by our definitions in these requirements) does not imply that fidelity relief is mandated for this modulation, which is
at almost the same absolute power levels for both the 64-QAM and QPSK bursts.
89

V.4 Examples Regarding Concurrent and Consecutive Changes in P
r_n
and P
load_min_set

load_min_set
load_min_set
to P
r_n
th
changes to P
r_n
load_min_set
. The CMTS should not issue a
change in P
load_min_set
r_n
until after also providing a sufficient reconfiguration time on
the n
th
channel. The CMTS should not issue a change in P
load_min_set
after commanding a prior change in P
load_min_set

until after also providing a global reconfiguration time for the first command. Also, the CMTS should not issue a
change in P
r_n
a) until after providing a global reconfiguration time following a command for a new value of
P
load_min_set
, and b) until after providing a sufficient reconfiguration time on the nth channel after issuing a previous
change in P
r_n
r_n
and/or P
load_min_set
to the CM
without a sufficient reconfiguration time for instituting the first command. When a concurrent new value of
P
load_min_set
and change in P
r_n
are commanded, the CM may wait to apply the change in P
r_n
at the next global
load_min_set
change at the first sufficient reconfiguration time of the nth channel; the value of P
load_min_set
which applies to the new
P
r_n
is the concurrently commanded P
load_min_set
value, and thus if the change to P
r_n
falls outside the Dynamic Range
Window of the old P
load_min_set
, then the CM has to wait for the global reconfiguration time to apply the change in P
r_n
.

89
11/17/11 CableLabs
159
In the following example, P
load_n
will be discussed along with P
r_n
, to simplify the example.
In this example, let P
load_min_set
be 15 dB, and P
load_1
is 16 dB and P
load_2
is 18 dB, and a new command is issued to
change P
load_min_set
to 14 dB, and concurrently to change P
load_1
to 14 dB. Now assume the opportunity exists to
change the power on Channel 1 but a global reconfiguration time is not provided until a bit later. Then the CM must
recognize that it should NOT discard the change in P
load_1
from 16 dB to 14 dB just because the new P
load_min_set
has
not had its opportunity to be applied. If the change in P
r_1
had been commanded BEFORE the change in P
load_min_set
,
then this change command to P
r_1
places P
r_1
outside of the Dynamic Range Window, and this command should be
ignored by the CM. However, since in the example the change to P
r_1
and P
load_min_set
was concurrent, the CM should
bypass the opportunity to change P
r_1
on Channel 1 when the quiet time on Channel 1 occurs, and the CM should
wait for the global reconfiguration time and then change both P
r_1
and P
load_min_set
.
If the change in P
r_1
had been only 1/2 dB, this would NOT have placed P
r_1
outside the starting Dynamic Range
Window, and the CM could have implemented the change in P
r_1
at the first opportunity, but the CM is allowed in
this case to wait and institute the change in P
r_1
at the same time as the change in P
load_min_set
, since the commands
were sent concurrently to the CM.
If the change in P
r_1
had been only 1/2 dB, and this change was commanded to the CM prior to the command to
change P
load_min_set
, then the change to P
r_1
must be implemented at the first reconfiguration opportunity on Channel 1.
If no such opportunity is provided and the change to P
load_min_set
is commanded, then the CM may ignore this
commanded change in P
load_min_set
, even though this command is in compliance with both the existing P
r_1
and the
pending P
r_1
; the CMTS should not issue such a command (the CMTS should wait for the CM to change P
r_1
before
sending a new P
load_min_set
). Note that the recommended response of the CM to such an improperly issued change in
P
load_min_set
is to continue with the change in P
r_1
at the provided Channel 1 reconfiguration time (as it is required to
do), and wait for the global reconfiguration time and implement (albeit inappropriately) the commanded P
load_min_set
.
Another circumstance in the event of the commanded change in P
load_min_set
prior to the opportunity for P
r_1
to change
is if the first subsequent opportunity for P
r_1
to change is also a global reconfiguration time; it IS allowed for the CM
to implement the change in P
r_1
at this opportunity, but NOT to change P
load_min_set
, since this was issued improperly
by the CMTS. However, it IS recommended that the CM hold on to this commanded new value of P
load_min_set
and
implement it at the NEXT global reconfiguration time.
Altering the example, if P
load_min_set
had been commanded to move from 15 dB to 17 dB with no changes in P
load_1
or
P
load_2
, this change is to be completely discarded since it places P
r_1
outside the Dynamic Range Window. If the
CMTS issues such a command, and before the next global reconfiguration time P
load_1
is commanded to change from
16 dB to 17 dB, it is still required that the CM ignore the P
load_min_set
change, since it was in violation of the Dynamic
Range Window constraints when it was issued.
160 CableLabs
11/17/11
Appendix VI Example Spurious Emissions Noise Power Limits with
Multiple Channels Bursting
This appendix provides examples illustrating the calculation of the noise power limits set by the spurious emissions
requirements for the upstream transmissions when more than one channel is bursting. The spurious emission
requirements are levied such that the noise power limits are relative to the channel transmit power; with multiple
channels bursting simultaneously, the absolute noise limits from each transmitted channel are added together to
produce the composite noise limit of the requirement. There are some conditions in the specification related to
channel transmit powers falling below certain levels relative to the other channel transmit powers, where the noise
limits are increased (relaxed). In these cases the noise power limit for the channel or channels which are transmitting
below the power range determined by the higher power channels corresponds to the transmit power for the low-
powered channel(s) if they were increased in power such as to fall just at the lower boundary of the power level
range rather than below the range.
dBmV, 45 dBmV, and 45 dBmV. The 41 dBmV power levels are raised (conceptually but not in actuality), for the
purposes of determining the spurious emissions limits, up to 53 dBmV 8 dB = 45 dBmV. If the modulation rate is
5120 kHz, and the noise limit is being calculated for a channel which is not adjacent to any of the three bursting
channels, then the noise limit corresponds to that which would occur if each bursting channel contributed noise
power -44 dB below its bursted signal power (or as in this example, below the artificially raised signal power for
channels falling below the ranges for full application of the requirements). Thus in this example, with all three
channels bursting, the noise limit for this test, in a non-adjacent transmit channel, is the sum of a) 53 dBmV - 44 dB
= 9 dBmV, b) 45 dBmV - 44 dB = 1 dBmV, and c) 45 dBmV - 44 dB = 1 dBmV. Converting to natural numbers
(rather than "dB" values) and adding the noise powers, the sum of 9 dBmV + 1 dBmV + 1 dBmV = 10.2 dBmV,
which serves as the noise power test limit for the spurious emissions requirements for this case.
As an example with four S-CDMA channels in the TCS, at transmit powers per channel of 53 dBmV, 41 dBmV, 41
dBmV, 49 dBmV, 45 dBmV, and 41 dBmV. The 41 dBmV power levels are raised (conceptually but not in
actuality), for the purposes of determining the spurious emissions limits, up to 53 dBmV 4 dB = 49 dBmV; and the
artificial 49 dBmV 4 dB = 45 dBmV; and the artificial 45 dBmV 4 dB = 41 dBmV. If the modulation rate is
5120 kHz, and the noise limit is being calculated for a channel which is not adjacent to any of the four bursting
channels, then the noise limit corresponds to that which would occur if each bursting channel contributed noise
power -44 dB below its bursted signal power (or as in this example, below the artificially raised signal power for
channels falling below the ranges for full application of the requirements). Thus in this example, with all four
channels bursting, the noise limit for this test, in a non-adjacent transmit channel, is the sum of a) 53 dBmV - 44 dB
= 9 dBmV, b) 49 dBmV - 44 dB = 5 dBmV, c) 45 dBmV - 44 dBmV = 1 dBmV, and d) 41 dBmV - 44 dB = -3
dBmV. Converting to natural numbers (rather than "dB" values) and adding the noise powers, the sum of 9 dBmV +
5 dBmV + 1 dBmV + (-3 dBmV) = 11.1 dBmV, which serves as the noise power test limit for the spurious emissions
requirements for this case.
11/17/11 CableLabs
161
Appendix VII Acknowledgements
On behalf of the cable industry and our member companies, CableLabs would like to thank the numerous individuals
that contributed to the development of this specification. In particular, we want to extend our sincere appreciation
and gratitude to the following PHY focus team members.

Contributor Company Affiliation Contributor Company Affiliation
Tom Ayers Broadlogic Jaehoo Lee ETRI
Zvi Bernstein ATI Rick Meller Cisco Systems
Liran Brecher Texas Instruments Jack Moran Motorola, Inc
L. Alberto Campos Cable Television Laboratories, Inc. Richard Prodan Broadcom Corporation
Phillip Chang Comcast Corp. Lakshmi Raman Cable Television Laboratories, Inc.
Bruce Currivan Broadcom Corporation Saifur Rahman Comcast Corp.
Kirk Erichsen Adelphia Vladimir Radionov Broadlogic
Jeff Finkelstein Cox Communications Carey Ritchie Microtune
Mike Grimwood ATI Hal Roberts BigBand Networks
Tom Kolze Broadcom Corporation Greg Taylor Cisco Systems
Bill Kostka Cable Television Laboratories, Inc. Robert Thompson Motorola, Inc.
David Hull Conexant Dan Torbet Arris International
Volker Leisse EuroCableLabs Daniel Wajcer Texas Instruments
Eungdon Lee ETRI

We would particularly like to thank Tom Kolze of Broadcom for his commitment and leadership as the PHY Focus
Team editor, Alberto Campos of CableLabs for all his efforts as Focus Team Lead, and all other participants who
served as liaisons to other focus teams.
Our sincere appreciation goes out to all the participating vendor companies and their contributing engineers, in
particular: Roger Fish of Broadcom Corporation; Xing Gao of Cisco Systems; Bill Kuiper of Microtune; Larry
O'Sullivan of Bigband Networks; Ali Somnez of Arris International; and Naor Goldman, Shaul Klein, and Davey
Weisberg of Texas Instruments.
Special thanks to Phillip Chang of Comcast and Scott Brown of Arris International for the in-depth specification
reviews; as well as to Scott Cummings, Lisa Denney, Victor Hou, and Niki Pantelias of Broadcom; Michael Patrick
of Motorola; Ariel Yagil of Texas Instruments; and Eduardo Cardona, Brian Hedstrom, Greg White, and Charlie
Bergren of CableLabs for their participation as invited specialists.
90

90
162 CableLabs
11/17/11
Appendix VIII Revision History
VIII.1 Engineering Changes incorporated into CM-SP-PHYv3.0-I02-061222
ECN ECN Date Summary
PHYv3.0-N-06.0301-2 11/1/2006 Update of Sections 6.2.20 and 6.2.20.1 regarding Ranging Offsets
PHYv3.0-N-06.0288-1 9/20/2006 MAP Processing Time for Non-MTC mode Operation

PHYv3.0-N-06.0328-2 12/13/2006
Terminology update to PHY Sections 3 (Terms and Definitions) and 4,
(Abbreviations and Acronyms)
PHYv3.0-N-06.0340-1 12/13/2006 PHY Omnibus EC #1 Clarification of "loading" and correction of minor typos
PHYv3.0-N-06.0366-2 1/24/2007
Introduce terms "Standard Upstream Frequency Range" & "Extended Upstream
Frequency Range"

PHYv3.0-N-07.0408-1 4/11/2007 Adjacent Channels Constraint number correction
PHYv3.0-N-07.0389-1 3/14/2007 Max 6 MHz average power of any carrier input to CM up to 1002 MHz

PHYv3.0-N-07.0465-1 6/13/2007 Minor Editorial Changes
PHYv3.0-N-07.0463-1 6/6/2007 Clarification of maximum average power input into CM in European technology option
PHYv3.0-N-07.0459-2 5/30/2007 Minor editorial alignments of European technology option with main text
PHYv3.0-N-07.0458-1 5/23/2007 Clarifying CM BER MUST Statement
PHYv3.0-N-07.0456-1 5/16/2007 Add MAY option for CM downstream tuning to 1002 MHz

PHYv3.0-N-07.0570-2 11/28/2007 Indirect dynamic range window changes
PHYv3.0-N-07.0598-1 1/16/2008 Correction to Burst Profile Tables
PHYv3.0-N-07.0599-1 1/16/2008 Editorial corrections to PHYv3.0

11/17/11 CableLabs
163
PHYv3.0-N-08.0611-3 3/19/2008 Dynamic Range Window with UCD Changing Phi
PHYv3.0-N-08.0619-1 3/26/2008 Fix Incorrect section number reference (ITU-T J.83-B)
PHYv3.0-N-08.0631-1 4/16/2008 Upstream Spurious Emissions in 85 to108 MHz

PHYv3.0-N-08.0701-2 12/10/08 PHYv3.0 Symbol Mapping Corrections and Edits
PHYv3.0-N-08.0702-2 12/10/08 Add CMCIv3.0 to references

VIII.8 Engineering Change incorporated into CM-SP-PHYv3.0-I09-101008
PHYv3.0-N-10.0942-5 9/15/10 Upstream power CM Capability in PHY3.0

PHYv3.0-N-11.1021-1 10/19/2011 Introduce Extended Upstream Frequency Range 5-85 MHz to European technology
option
PHYv3.0-N-11.1020-2 10/19/2011 Align DS frequency range across European technology options in DOCSIS
specifications

Data Over Cable Service Interface Specifications Physical Layer Specificaction

Загружено:

Сведения о документе

Исходное описание:

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Data Over Cable Service Interface Specifications Physical Layer Specificaction

Загружено:

Авторское право:

Доступные форматы

Data Over Cable Service Interface Specifications

family of specifications developed by Cable Television Laboratories

Вам также может понравиться